Learning Archives - Business Intelligence Info

Tag Archives: Learning

Yann LeCun and Yoshua Bengio: Self-supervised learning is the key to human-level intelligence

May 3, 2020 Big Data

Self-supervised learning could lead to the creation of AI that’s more human-like in its reasoning, according to Turing Award winners Yoshua Bengio and Yann LeCun. Bengio, director at the Montreal Institute for Learning Algorithms, and LeCun, Facebook VP and chief AI scientist, spoke candidly about this and other research trends during a session at the International Conference on Learning Representation (ICLR) 2020, which took place online.

Supervised learning entails training an AI model on a labeled data set, and LeCun thinks it’ll play a diminishing role as self-supervised learning comes into wider use. Instead of relying on annotations, self-supervised learning algorithms generate labels from data by exposing relationships among the data’s parts, a step believed to be critical to achieving human-level intelligence.

“Most of what we learn as humans and most of what animals learn is in a self-supervised mode, not a reinforcement mode. It’s basically observing the world and interacting with it a little bit, mostly by observation in a test-independent way,” said LeCun. “This is the type of [learning] that we don’t know how to reproduce with machines.”

But uncertainty is a major barrier standing in the way of self-supervised learning’s success.

VB Transform 2020 Online – July 15-17, 2020: Join leading AI executives at VentureBeat’s AI event of the year. Register today and save 30% off digital access passes.

Distributions are tables of values — they link every possible value of a variable to the probability the value could occur. They represent uncertainty perfectly well where the variables are discrete, which is why architectures like Google’s BERT are so successful. Unfortunately, researchers haven’t yet discovered a way to usefully represent distributions where the variables are continuous — i.e., where they can be obtained only by measuring.

LeCun notes that one solution to the continuous distribution problem is energy-based models, which learn the mathematical elements of a data set and try to generate similar data sets. Historically, this form of generative modeling has been difficult to apply practically, but recent research suggests it can be adapted to scale across complex topologies.

For his part, Bengio believes AI has much to gain from the field of neuroscience, particularly its explorations of consciousness and conscious processing. (It goes both ways — some neuroscientists are using convolutional neural networks, a type of AI algorithm well-suited to image classification, as a model of the visual system’s ventral stream.) Bengio predicts that new studies will elucidate the way high-level semantic variables connect with how the brain processes information, including visual information. These variables are the kinds of things that humans communicate using language, and they could lead to an entirely new generation of deep learning models.

“There’s a lot of progress that could be achieved by bringing together things like grounded language learning, where we’re jointly trying to understand a model of the world and how high-level concepts are related to each other. This is a kind of joint distribution,” said Bengio. “I believe that human conscious processing is exploiting assumptions about how the world might change, which can be conveniently implemented as a high-level representation. Those changes can be explained by interventions, or … the explanation for what is changing — what we can see for ourselves because we come up with a sentence that explains the change.”

Another missing piece in the human-level intelligence puzzle is background knowledge. As LeCun explained, most humans can learn to drive a car in 30 hours because they’ve intuited a physical model about how the car behaves. By contrast, the reinforcement learning models deployed on today’s autonomous cars started from zero — they had to make thousands of mistakes before figuring out which decisions weren’t harmful.

“Obviously, we need to be able to learn models of the world, and that’s the whole reason for self-supervised learning — running predictive models of the world that would allow systems to learn really quickly by using this model,” said LeCun. “Conceptually, it’s fairly simple — except in uncertain environments where we can’t predict entirely.”

LeCun argues that even self-supervised learning and learnings from neurobiology won’t be enough to achieve artificial general intelligence (AGI), or the hypothetical intelligence of a machine with the capacity to understand or learn from any task. That’s because intelligence — even human intelligence — is very specialized, he says. “AGI does not exist — there is no such thing as general intelligence,” said LeCun. “We can talk about rat-level intelligence, cat-level intelligence, dog-level intelligence, or human-level intelligence, but not artificial general intelligence.”

But Bengio believes that eventually, machines will gain the ability to acquire all kinds of knowledge about the world without having to experience it, likely in the form of verbalizable knowledge.

“I think that’s a big advantage for humans, for example, or with respect to other animals,” he said. “Deep learning is scaling in a beautiful way, and that’s one of its greatest strengths, but I think that culture is a huge reason why we’re so intelligent and able to solve problems in the world … For AI to be useful in the real world, we’ll need to have machines that not just translate, but that actually understand natural language.”

Let’s block ads! (Why?)

Big Data – VentureBeat

UC Berkeley researchers open-source RAD to improve any reinforcement learning algorithm

May 2, 2020 Big Data

A group of University of California, Berkeley researchers this week open-sourced Reinforcement Learning with Augmented Data (RAD). In an accompanying paper, the authors say this module can improve any existing reinforcement learning algorithm and that RAD achieves better compute and data efficiency than Google AI’s PlaNet, as well as recently released cutting-edge algorithms like DeepMind’s Dreamer and SLAC from UC Berkeley and DeepMind.

RAD achieves state-of-the-art results on common benchmarks and matches or beats every baseline in terms of performance and data efficiency across 15 DeepMind control environments, the researchers say. It does this in part by applying data augmentations for visual observations. Coauthors of the paper on RAD include Michael “Misha” Laskin, Kimin Lee, and Berkeley AI Research codirector and Covariant founder Pieter Abbeel.

RAD was released Thursday on preprint repository arXiv. Data augmentation has been important to advances in convolutional neural networks (CNN) for challenges like robotic grasping and achieving human-level performance in games like Go.

“For the first time, we show that data augmentations alone can significantly improve the data-efficiency and generalization of RL methods operating from pixels, without any changes to the underlying RL algorithm, on the DeepMind Control Suite and the OpenAI ProcGen benchmarks, respectively,” the paper reads. “By using multiple augmented views of the same data point as input, CNNs are forced to learn consistencies in their internal representations. This results in a visual representation that improves generalization, data-efficiency, and transfer learning.”

VB Transform 2020 Online – July 15-17, 2020: Join leading AI executives at VentureBeat’s AI event of the year. Register today and save 30% off digital access passes.

Data augmentation techniques increase diversity in training data sets without collecting new data. “We find that data diversity alone can make agents focus on meaningful information from high-dimensional observations without any changes to the reinforcement learning method,” the authors note.

It’s been a busy week for the machine learning subfield of reinforcement learning.

Earlier this week, NYU researchers released work on arXiv that applies data augmentation they say also achieves state-of-the-art results on the DeepMind control suite.

And at the entirely digital International Conference on Learning Representations (ICLR) this week, Google AI researchers introduced methods for measuring the reliability of reinforcement learning algorithms, and Huawei AI researchers introduced Adversarial AutoAugment for improving data augmentation policy.

Abbeel also coauthored a number of reinforcement algorithm papers at ICLR, including HiPPO for training several levels of reinforcement learning algorithms at once and a paper on reinforcement learning and policy optimization that touches on data augmentation.

In a different series of developments, earlier this week Salesforce released the AI Economist, reinforcement learning the company claims is able to create optimal tax policies.

Let’s block ads! (Why?)

Big Data – VentureBeat

Health care organizations use Nvidia’s Clara federated learning to improve mammogram analysis AI

April 16, 2020 Big Data

A group of health care organizations in the United States and Brazil have used federated learning to improve AI mammography classification models for recognizing tumors and breast tissue density assessment. An Nvidia spokesperson declined to share specific performance details, but said a combined effort led to better classification model performance than models made by any of the individual health care organizations. Impacting 2.1 million people a year, according to the World Health Organization, breast cancer is the most frequently occurring cancer in women around the world.

Federated learning is a way to create AI models from data in multiple locations without a need to store all the data in the same place. Models are trained at their source and learnings are shared with a single generalized model without the transfer of data. Apple and Google use federated learning for keyboard personalization today, but in health care, federated learning can combine anonymized data to improve machine intelligence geared toward saving human lives.

“Federated learning presents an opportunity for health care organizations worldwide to work together without compromising on the data security of patient records,” AI and health researcher Jayashree Kalpathy-Cramer said in an Nvidia blog post today sharing the news. “With this methodology, we can collectively raise the bar for AI tools in medicine.”

The work is the result of a collaboration that took place between January and March and pools resources from the American College of Radiology, Brazilian imaging center Diagnosticos da America, Partners HealthCare, The Ohio State University, and Stanford Medicine. The federated learning model was trained using more than 130,000 images from 33,000 mammography studies, Nvidia VP of health care Kimberly Powell told VentureBeat. Hospitals worked together using the Clara Federated Learning SDK.

This is the second such combined demonstration of federated learning by health care researchers using Nvidia’s Clara Medical Imaging platform. Last fall, Nvidia and Kings College London worked together on a federated learning neural network for brain tumor segmentation.

In other Clara news, Powell said Nvidia is working with collaborators to release coronavirus-related models trained with federated learning through the Clara Imaging Software in the future. Last week, Nvidia launched Clara Deploy 5.4 for multi-domain pipelines on a unified GPU infrastructure.

Let’s block ads! (Why?)

Big Data – VentureBeat

Lendio raises $55 million to surface small business loans using machine learning

March 2, 2020 Big Data

Lendio, a company that provides a platform that taps algorithms to match customers with lenders and secure loans, announced that it raised $ 55 million. CEO Brock Blake says the influx of funds will bolster Lendio’s marketplace while enabling it to introduce new bookkeeping and lender services functions.

Lendio claims to make it easier for small- and medium-sized business owners to secure loans. Using machine learning, it can streamline the whole process, which usually includes comparing lenders and juggling things like credit scores, collateral, and fees

“This … round is a significant milestone in Lendio’s mission to fuel the American dream as a financial matchmaker that brings high-tech, high-touch capital to the small business economy,” added Blake, who cofounded the South Jordon, Utah-based company in 2011 with Levi King and Trent Miskin. “With these funds, we are strongly positioned to grow our existing platform as a trusted loan facilitator that supports both lenders and borrowers, while building out a range of new integrated lending services that get the right loans into the right hands at the right time.”

On Lendio’s website, businesses fill out a fee- and obligation-free application that takes about 15 minutes and doesn’t impact their credit score. Then, they’re able to explore business line of credit, small business loan, short-term loan, business-term loan, merchant cash advance, business credit card, equipment financing, commercial mortgage, accounts receivable financing, startup loan, and business acquisition offers from over 75 partners including American Express, Chase, PayPal, LendingClub, Kabbage, Comcast Business, Staples, and Funding Circle.

The machine learning algorithms underpinning Lendio’s marketplace take into account factors like offer likelihood (i.e., whether the lender will give the applicant an offer), funding likelihood (whether the applicant is likely to accept an offer), and the applicant’s product market. According to Brock, they’re trained on over 140 dimensions about the borrowers and their business and more than 800,000 possible deals, enabling Lendio to find financing for more than 65% of applicants compared with the 20% who successfully obtain bank loans. “This helps capture the ideal balance between the lender’s funding target and what works well for the borrower, and deep integrations with lenders allow for programmatic and instant submissions,” he told VentureBeat via email.

Lendio says it’s funded $ 1.4 billion in capital across 73,000 loans, which the company asserts makes its marketplace the largest of its kind in the U.S. Over the past two years, year-over-year growth averaged 75%, and the number of loans facilitated grew to over 100,000 (totaling $ 2 billion),

Brock says the financing from this latest round will also be used to grow Lendio’s Sunrise product, a money management dashboard that helps customers visualize their cash flow and finances. It supports customized invoices and lets users link both bank accounts and credit cards, and it ships with templates for profit and loss, balance sheet, tax summary, customer statement, and accounts receivable reports. Business owners can invite accountants to access transactions and financial reports, who can add sales tax to invoices based on customers’ locations, and they benefit from backups across multiple data centers over encrypted connections.

Additionally, Lendio plans to enhance its lender services division, which provides banks, credit unions, and online lenders access to a custom, white-labeled online app that lets them accept loan applications online.

About $ 31 million of Lendio’s latest raise was equity, led by Mercato Partners’ Traverse Fund, while $ 24 million was a debt facility from Signature Bank. The company has raised over $ 100 million in venture capital to date, and it employs over 300 people according to LinkedIn data.

Let’s block ads! (Why?)

Big Data – VentureBeat

SQL Server Machine Learning 2019: Working with Security Changes

February 24, 2020 BI News and Info

I have a confession to make. Why, in my last article about shortest_path in SQL Server 2019, have I used Gephi in order to illustrate the relationships, instead of using a script in R for the same purpose and demonstrate Machine Learning Services as well?

The initial plan was to use an R script; however, the R script which works perfectly in SQL Server 2017 doesn’t work in SQL Server 2019.

This is the original R script I had planned to use:

EXEC Sp_execute_external_script

@language = N‘R’,

@script = N‘ require(igraph)

g <- graph.data.frame(graphdf)

V(g)$ label.cex <- 2

png(filename = “c:\R\plot1.png”, height = 1200, width = 1200, res = 100);

plot(g, vertex.label.family = “sans”, vertex.size = 40)

dev.off() ‘,

@input_data_1 = N‘select LikeMember.MemberName as LikeMember,

LikedMember.MemberName as LikedMember

from dbo.ForumMembers as LikeMember, dbo.ForumMembers as LikedMember,

Likes

where Match(LikeMember-(Likes)->LikedMember)’,

@input_data_1_name = N‘graphdf’

The purpose of the script was only to create an image file with the relationships stored in graph tables, but in SQL Server 2019, it results in an access denied message.

If you would like to reproduce this demonstration, you can use this script file to create the database in a SQL Server 2017. You must also install the R packages needed by the script. There are many ways to do it. My favourite is this:

Open windows explorer
Browse to the SQL Server R folder, usually is this one: C:\Program Files\Microsoft SQL Server\MSSQL15.MSSQLSERVER\R_SERVICES\bin
Execute the R.EXE application
Execute the instruction install.packages(“igraph”) , where “igraph” is the name of the package to be installed

When you try and run the above R script. It will work well in SQL Server 2017, and fail in SQL Server 2019 with the error message you can see in the image below:

a screenshot of a cell phone description automati 14 SQL Server Machine Learning 2019: Working with Security Changes

During PASS Summit, I was able to ask questions to the Microsoft employees in the Data Clinic, and they explained the differences between the two versions. The explanation hasn’t solved all the problems, but I was able to find a solution and create a workaround for this script.

While researching the workaround, I saw many examples of R scripts manipulating files and all these examples may fail in SQL Server 2019. Highlighting this problem, solution, and workaround in this article will be important to everyone using SQL Server Machine Learning and planning to migrate to SQL Server 2019.

First, I’ll review how Machine Learning in SQL Server works: it involves the execution of an external script, in R or Python, but has a flexible structure, you could create your own external language. The execution itself is started by a service called SQL Server Launchpad, isolating the execution from SQL Server itself. The image below (from What is the Machine Learning Server) illustrates how the execution architecture works, although this article will not go so deep into the architecture.

a screenshot of a cell phone description automati 15 SQL Server Machine Learning 2019: Working with Security Changes

Launchpad in SQL Server 2017

In SQL Server 2017 the Launchpad starts processes for each execution. The processes need an identity, so SQL Server dynamically creates users inside the group SQLRUserGroup. In the following image, you can see two groups, SQLRUserGroup, for SQL Server 2019 and SQLRUserGroupSQL2017 for SQL Server 2017, since I have both in the same machine.

The difference in the name is due to the instance name; SQL Server 2019 is the default instance in my machine. Due to that, its user group name has only the core name, while SQL Server 2017 is in an instance called SQL2017, so the instance name is attached to the core group name.

The processes have access to the machine resources, limited by the permissions given to the SQLRUserGroup. Since SQL Server 2017 creates many users, you shouldn’t grant permissions directly to the users, only to the group.

In order to be able to manipulate files in the file system, you only need to give the correct permissions to the group SQLRUserGroup. The image below shows the group and users for SQL Server 2017

a screenshot of a cell phone description automati 16 SQL Server Machine Learning 2019: Working with Security Changes

Launchpad in SQL Server 2019

In SQL Server 2019, the Launchpad was improved. Instead of using processes for each execution, it uses app containers. The app containers need only a single identity, so they use the same identity as the Launchpad service.

While the 2017 version, the SQLRUserGroup, had many users; in the 2019 version, it has only one, the Launchpad service account. The image below shows the group for SQL Server 2019.

a screenshot of a cell phone description automati 17 SQL Server Machine Learning 2019: Working with Security Changes

Nowadays, when talking about containers immediately, we think about Docker, but not in this case. The app containers I’m mentioning here are a sandbox set of APIs introduced in Windows 8. These allow any application to create app containers and start an execution inside an app container.

The app containers work as isolated virtual machines (but much more lightweight, of course). This means that you can’t access the file system of the host (your SQL Server), neither for reading or to save files. You save any files inside the app container, and they are destroyed after the app container is released. In fact, the container is redirected to save files in a temporary folder inside the host hard drive. This increases the security, no doubt about that, but any script dealing with the file system may need changes to work in SQL Server 2019.

These app containers are created by Launchpad using the windows Sandbox API. The Launchpad doesn’t offer any way to customize the app containers creation, which could solve the problem.

You can learn more about app containers with these two links:

https://techcommunity.microsoft.com/t5/Windows-Kernel-Internals/Windows-Sandbox/ba-p/301849
https://www.malwaretech.com/2015/09/advanced-desktop-application-sandboxing.html

Proving this Concept

In order to determine this concept, you can use a script in R to access the disk and list the files. You can execute the script below in both SQL Server versions, 2017 and 2019, and compare the results. In SQL Server 2017, the script will have access to the host disk. In SQL Server 2019, the script will have access only to a fake disk, which is, in fact, a temporary folder in the host disk and will be deleted after its use by the app container.

EXEC Sp_execute_external_script

@language = N‘R’,

@script = N‘data <- list.files()

data2 <- data.frame(data)’,

@output_data_1_name = N‘data2′

EXEC Sp_execute_external_script

@language = N‘R’,

@script = N‘data <- list.dirs(“..\”)

data2 <- data.frame(data)’,

@output_data_1_name = N‘data2′

EXEC Sp_execute_external_script

@language = N‘R’,

@script = N‘data <- list.dirs(“..\..”)

data2 <- data.frame(data)’,

@output_data_1_name = N‘data2′

The image below the script shows the result in SQL Server 2019.

Solutions

The app containers are created during the SQL Server Machine Learning services setup. They are objects inside the Windows local directory, and such as all the objects in the local directory, they have a unique SID to identify the app container.

You can define file system permissions directly for the SIDs. The problem is: How to identify the SID’s of the app containers since they are not listed as user or groups.

Analyse two possible solutions:

Easy and tempting: You can give permission to the object called All Application Containers. The R scripts will have the file system permission; however, any other app container eventually running on the same machine will have the file system permission as well.
Secure: You can identify the SIDs of the app containers installed by SQL Server and give permissions directly to them. In this way, only the app containers used by the SQL Server Machine Learning Services will receive these permissions

You can test both solutions using a straightforward script. First, create a folder called C:\testFolder and copy some files, any files, to the folder. Try to list the files in this folder using the script below.

EXEC Sp_execute_external_script

@language = N‘R’,

@script = N‘data <- list.files(“c:\testFolder”)

data2 <- data.frame(data)’,

@output_data_1_name = N‘data2′

All Application Containers

The All Application Container object has a fixed SID in the local directory, which is S-1-15-2-1 . You can use the application icacls to grant permission

Execute the R script above against the 2019 instance using SSMS. The script will return no result
Press Win+R and type CMD to open a command prompt as administrator
Type the following instruction and press ENTER

icacls c:\testFolder /grant *S-1-15-2-1:(OI)(CI)F /t

Using Windows Explorer, right-click the testFolder and select Properties in the context menu
In the Properties window, click the Security tab. You will be able to confirm the permission, like the image below.

Open SQL Server Configuration Manager
On the left side of SQL Server Configuration Manager, select SQL Server Services
In the right-side of SQL Server Configuration Manager, right-click the Launchpad service and select Restart item in the context menu
Execute the same R script again in SSMS. This time it will work and list the files in the folder.

a screenshot of a cell phone description automati 18 SQL Server Machine Learning 2019: Working with Security Changes

Delete the testFolder folder so that you can continue with the other solution below

Using the App Containers SID

The challenge you will face when giving permissions to the App Containers SID is to discover which are the App Containers SID. There is not a direct solution for that, although, after knowing where to look, it becomes easy. SQL Server installs firewall rules in Windows Firewall in order to forbid app containers from making external contact to the network. These firewall rules are created precisely to block the app containers SID from using the network, so you can identify these SIDs by analysing these firewall rules.

If you look at the firewall rules, you can see that the app containers are blocked.

word image 81 SQL Server Machine Learning 2019: Working with Security Changes

The firewall UI doesn’t provide the SIDs, however. You need to dig deeper using PowerShell to retrieve the SIDs from the firewall rules.

The PowerShell command is a combination of the cmdlet Get-NetFirewallRule, to retrieve the firewall rules and the cmdlet Get-NetFirewallApplicationFilter in order to retrieve, from each rule, the app container filter information.

The complete PowerShell command line is this:

(Get-NetFirewallRule | Where-Object { ($ _.Direction -eq “Outbound”) -and ($ _.DisplayName -like “*appcontainer*”)}) | %{ ($ _ | Get-NetFirewallApplicationFilter) | %{Write-Output $ _.Package } }

Analyse this command line in more detail:

Get-NetFirewallRule will retrieve all the firewall rules in Windows Firewall
Where-Object filters the firewall rules, retrieving only the outbound rules which contains ‘appcontainer’ in the name
The expression %{ } is a shortcut to foreach-object so that the instructions will be executed for each firewall rule
Get-NetFirewallApplicationFilter is executed for each firewall rule, having the firewall rule ($ _ ) as a parameter
The expression % { } is used again to run one line for each application filter found
The Write-Output shows the value of the Package property on each application filter

a black and blue text description automatically g 1 SQL Server Machine Learning 2019: Working with Security Changes

Having discovered the SIDs, you need to set the permission to the folder. The instruction will look like the one below for each SID.

ICACLS “C:\testFolder” /grant “*S-1-15-2-1853514363-3573294594-1452771951-225645021-3819702349-824866239-3302992986″:(OI)(CI)F /t

You could copy and paste all the sids returned to create a batch file to set the permissions, but PowerShell can do all of this you as well. The following script prompts you for a file path, the one where you want Machine Learning to have access, and sets the permissions:

Param (

[Parameter(

Mandatory)]

[string]

$ FolderName

)

Get-NetFirewallRule |

Where-Object {

$ _.Direction -eq “Outbound” -and

$ _.DisplayName -like “*appcontainer*”

} |

ForEach-Object {

$ _ |

Get-NetFirewallApplicationFilter |

ForEach-Object {

$ Acl = Get-Acl $ FolderName

$ Sec= New-Object System.Security.Principal.SecurityIdentifier($ _.Package)

$ AccessRule = New-Object System.Security.AccessControl.FileSystemAccessRule(

$ Sec,

“FullControl”,

(1 -bor 2),

“Allow”

)

$ Acl.SetAccessRule($ AccessRule)

$ Acl |

Set-Acl $ FolderName

} # end 2nd FE-O

} # end 1st FE-O

The script is using COM objects System.Security.Principal.SecurityIdentifier and System.Security.AccessControl.FileSystemAccessRule in order to set the access rule on the folder.

After running the script, take a look at the security properties of the folder, and the R script should now work from SQL Server 2019.

word image 82 SQL Server Machine Learning 2019: Working with Security Changes

Workaround with a FileTable

Although I have identified two solutions, it’s interesting to notice Microsoft is moving SQL Server Machine Learning towards a more secure environment, avoiding network and file system access. Considering this, you may like to use an alternative solution.

There are some options, all based on returning the image as an output parameter to SQL Server. Here is one of the options.

A filetable can map a disk folder as a table in SQL Server. This a very good solution when you need to deal with files. The script in R will see the table as a regular table. While reading or inserting into the table, the script will be reading and inserting from/to the disk.

It’s a very good solution for a permanent environment, however, it has some problems:

You can’t define the final location of each file using SQL Server configuration. The filetable will be exposed from the server as a sharing folder, not a disk folder, so there is a slight difference in relation to the initial planning.
Filetable is not supported in Azure PaaS cloud solutions, so you will be tied to IaaS solutions, which means creating and maintaining a virtual machine in Azure, or an on-premise environment.

The Filetable technology was created in SQL Server 2016, based on the filestream technology which was built on the previous version.

Enabling Filetable in SQL Server

A filetable solution needs some administrative care; for this reason, this solution can only be used after enabling it on a server configuration level. Follow these steps to enable this solution:

Open SQL Server Configuration Manager
On the left side of SQL Server Configuration Manager, select SQL Server Services
On the right side, double click the SQL Server service you are using
On the SQL Server properties window, select the FileStream tab

a screenshot of a cell phone description automati 19 SQL Server Machine Learning 2019: Working with Security Changes

In the filestream tab, check the box Enable filestream for Transact-SQL Access
In the filestream tab, check the box Enable filestream for I/O Access
The textbox Windows Share Name will be filled with the default value MSSQLSERVER. You can change it, but leave this default for now
In the filestream tab, check the box Allow remote clients to access filestream data.
Click the Ok button in the SQL Server Properties window
In the right-side of SQL Server Configuration Manager, right-click the SQL Server service and select Restart item in the context menu

a screenshot of a cell phone description automati 20 SQL Server Machine Learning 2019: Working with Security Changes

In SSMS, change the FileStream access level configuration using the following code:

EXEC sp_configure ‘filestream access level’, 2

RECONFIGURE

Value 1 enables filestream only for T-SQL access, which is not enough for filetable. Value 2 enables filestream for T-SQL and I/O access.

Preparing the Database to Support Filetable

In order to support filetable, you need to make some changes to the database. You need to create a filegroup to contain the filestream data, configure a special file, which will point to a folder and change some database settings.

Here is the T-SQL to execute for the database. Be sure to delete the c:\R folder before running this code:

ALTER DATABASE graphdemo ADD filegroup filestreamgroup

CONTAINS filestream

ALTER DATABASE graphdemo ADD FILE (NAME=fs, filename=‘c:\R’)

TO filegroup filestreamgroup

In SSMS, Object Explorer, right-click the GraphDemo database and select Properties in the context menu

In the database properties window, set the property Filestream Directory Name to RFiles. During further steps, you will better understand what this means.
In the database properties window, set the property Filestream Non-Transaction access to Full

In the database properties window, click the Ok button
Confirm the close of database connections clicking the Ok button in the dialog box that will appear

a screenshot of a cell phone description automati 21 SQL Server Machine Learning 2019: Working with Security Changes

Create the filetable using the following T-SQL:

CREATE TABLE images AS filetable WITH

(

filetable_directory = ‘Images’

);

The filetable you created, Images, is linked to a folder. Which folder? If you check the path c:\R, which was used to configure the database file, you will find some files and folders, but it’s not easy to understand them. So, what’s missing?

Follow these steps to see:

Click win+R
Type \localhost and click Ok You will find the share MSSQLSERVER, that’s the name you configured in SQL Server Configuration Manager. Now that you know how the name is used, you can choose a better name if needed.

Open the share MSSQLSERVER You will find a folder called RFiles. That’s the configuration made for the database. Each database configured to use filestream will have a share inside \Localhost\MSSQLSERVER

a screenshot of a cell phone description automati 22 SQL Server Machine Learning 2019: Working with Security Changes

Open the folder Rfiles You will find another folder called Images, the name used for the filetable. Each filetable in the database GraphDemo will have its own folder inside \localhost\MSSQLSERVER\RFILES

a screenshot of a cell phone description automati 23 SQL Server Machine Learning 2019: Working with Security Changes

Open the folder Images. The folder is empty.
In SSMS, execute the following query:

select * from Images

The filetable is empty.
Copy a file, any file, to the Images folder, \localhost\MSSQLServer\RFILES\Images
Execute the select again in SSMS:

select * from Images

This time the table has one row, exactly the row for the new file you copied. The filetable is linked with the folder. So, if you insert a record inside the filetable, the record will be a file available in the folder, in this way you can save to disk the image generated by the R script

New R Script for Filetable

After preparing the filetable environment, the new script will be this one:

DECLARE @img VARBINARY(max)

EXEC Sp_execute_external_script

@language = N‘R’,

@script = N‘ require(igraph)

require(hexView)

g <- graph.data.frame(graphdf)

V(g)$ label.cex <- 2 png(filename = “plot1.png”, height = 1200, width = 1200, res = 100);

plot(g, vertex.label.family = “sans”, vertex.size = 40)

dev.off()

img <- readRaw(“plot1.png”)

imageContent <- img$ fileRaw’,

@input_data_1 = N‘select LikeMember.MemberName as LikeMember,

LikedMember.MemberName as LikedMember

from dbo.ForumMembers as LikeMember, dbo.ForumMembers as LikedMember,

Likes

where Match(LikeMember-(Likes)->LikedMember)’,

@input_data_1_name = N‘graphdf’,

@params = N‘@imageContent varbinary(max) OUTPUT’

@imageContent = @img output;

INSERT INTO images (NAME, file_stream)

VALUES (‘test.png’, @img)

The script is very similar to the initial one; however, this time, it results in an output variable with the image.

Here are the differences between this script and the original:

A varbinary(max) variable which will be used as an output
A new R library, hexView, must be installed in order to do additional manipulations in the file (yes, you will still use a file)
After saving the plot to disk, it’s read back, reading the raw file content
The variable where you stored the raw image (imageContent) is defined as a parameter using the @params argument of the sp_execute_external_script
You need to link this script variable with a variable outside the script to get the content
Finally, you insert the variable into the filetable

The result of the execution of the script above will be a new file in the folder, you can check by accessing \localhost\MSSQLSERVER\RFILES\Images

a screenshot of a cell phone description automati 25 SQL Server Machine Learning 2019: Working with Security Changes

Conclusion

During my research on the web, I saw many examples of R scripts dealing with the file system. These scripts which are running nowadays in SQL Server 2017 will break in SQL Server 2019 unless one of the changes explained in this article is applied.

You can choose to fix the file system permission or start to avoid direct file system access from the R script. I believe Microsoft will provide a fix for this problem soon, what probably will be an easier way to give permissions to the app container SIDs

Let’s block ads! (Why?)

SQL – Simple Talk

Researchers propose data set to measure few-shot learning performance

February 20, 2020 Big Data

A plethora of AI models have been tailored to tackle few-shot classification, which refers to learning a classifier for new classes given only a few examples (an ability humans naturally possess). Improving on it could lead to more efficient algorithms capable of expanding their knowledge without requiring large labeled data sets, but to date, many of the procedures and corpora used to assess progress here are lacking.

That’s why researchers at Google AI, the University of California, Berkeley, and the University of Toronto propose in a preprint paper a benchmark for training and evaluating large-scale, diverse, and more “realistic” few-shot classification models. They say it improves upon previous approaches by incorporating multiple data sets of “diverse” distributions and by introducing realistic class imbalance, which they say allows the testing of robustness across a spectrum from low-shot learning onward.

The work was published in May 2019, but it was recently accepted to the International Conference on Learning Representations (ICLR) that’ll be held in Addis Ababa, Ethiopia in April.

As the team explains, as opposed to synthetic environments, real-life learning experiences are heterogeneous in that they vary by the number of classes and the examples per class. They also measure only within-corpus generalization, and they ignore the relationships between classes when forming episodes — i.e., the coarse-grained classification of dogs and chairs may present different difficulties than the fine-grained classification of dog breeds. (An “episode” encompasses states that come between an initial-state and a terminal-state, such as a game of Chess.)

By contrast, the researchers’ data set — the Meta-Dataset — leverages data from 10 different corpora, which span a variety of visual concepts natural and human-made and vary in the specificity of the class definition. Two are reserved for evaluation, meaning that no two classes from them participate in the training set, while the remaining ones contribute some classes to each of the training, validation, and test splits of classes.

Meta-Dataset separately employs an algorithm for sampling episodes, which aims to yield imbalanced episodes of variable shots (class precision) and ways (accuracy). A prescribed number of examples of each chosen class are chosen uniformly at random to populate the support and query sets.

In experiments, the team trained meta-learning models via training episodes sampled using the same algorithm as used for Meta-Dataset’s evaluation episodes. They say that, tested against Meta-Dataset, the models generally didn’t improve when provided multiple data sources and that they didn’t benefit from meta-learning across the data sets. Moreover, they report that the models weren’t robust to the amount of data in test episodes; rather, each excelled in a different part of the spectrum.

“We believe that our exploration of various models on Meta-Dataset has uncovered interesting directions for future work pertaining to meta-learning across heterogeneous data,” wrote the coauthors, who added that addressing the uncovered shortcomings constitutes an important research goal. “[I]t remains unclear what is the best strategy for creating training episodes, the most appropriate validation creation and the most appropriate initialization.”

Let’s block ads! (Why?)

Big Data – VentureBeat

AI Weekly: Machine learning could lead cybersecurity into uncharted territory

February 15, 2020 Big Data

Once a quarter, VentureBeat publishes a special issue to take an in-depth look at trends of great importance. This week, we launched issue two, examining AI and security. Across a spectrum of stories, the VentureBeat editorial team took a close look at some of the most important ways AI and security are colliding today. It’s a shift with high costs for individuals, businesses, cities, and critical infrastructure targets — data breaches alone are expected to cost more than $ 5 trillion by 2024 — and high stakes.

Throughout the stories, you may find a theme that AI does not appear to be used much in cyberattacks today. However, cybersecurity companies increasingly rely on AI to identify threats and sift through data to defend targets.

Security threats are evolving to include adversarial attacks against AI systems; more expensive ransomware targeting cities, hospitals, and public-facing institutions; misinformation and spear phishing attacks that can be spread by bots in social media; and deepfakes and synthetic media have the potential to become security vulnerabilities.

In the cover story, European correspondent Chris O’Brien dove into how the spread of AI in security can lead to less human agency in the decision-making process, with malware evolving to adapt and adjust to security firm defense tactics in real time. Should costs and consequences of security vulnerabilities increase, ceding autonomy to intelligent machines could begin to seem like the only right choice.

We also heard from security experts like McAfee CTO Steve Grobman, F-Secure’s Mikko Hypponen, and Malwarebytes Lab director Adam Kujawa, who talked about the difference between phishing and spear phishing, addressed an anticipated rise in personalized spear phishing attacks ahead, and spoke generally to the fears — unfounded and not — around AI in cybersecurity.

VentureBeat staff writer Paul Sawers took a look at how AI could be used to reduce the massive job shortage in the cybersecurity sector, while Jeremy Horwitz explored how cameras in cars and home security systems equipped with AI will impact the future of surveillance and privacy.

AI editor Seth Colaner examines how security and AI can seem heartless and inhuman but still relies heavily on people, who are still a critical factor in security, both as defenders and targets. Human susceptibility is still a big part of why organizations become soft targets, and education around how to properly guard against attacks can lead to better protection.

We don’t know yet the extent to which those carrying out attacks will come to rely on AI systems. And we don’t know yet if open source AI opened Pandora’s box, or to what extent AI might increase threat levels. One thing we do know is that cybercriminals don’t appear to need AI to be successful today.

I’ll leave it to you to read the special issue and draw your own conclusions, but one quote worth remembering comes from Shuman Ghosemajumder, formerly known as the “click fraud czar” at Google and now CTO at Shape Security, in Sawers’ article. “[Good actors and bad actors] are both automating as much as they can, building up DevOps infrastructure and utilizing AI techniques to try to outsmart the other,” he said. “It’s an endless cat-and-mouse game, and it’s only going to incorporate more AI approaches on both sides over time.”

For AI coverage, send news tips to Khari Johnson and Kyle Wiggers and AI editor Seth Colaner — and be sure to subscribe to the AI Weekly newsletter and bookmark our AI Channel.

Thanks for reading,

Khari Johnson

Senior AI Staff Writer

Let’s block ads! (Why?)

Big Data – VentureBeat

7 Questions on How to Use Machine Learning for Anomaly Detection

January 29, 2020 TIBCO Spotfire

Reading Time: 3 minutes

Asking questions is one of the best ways to learn. But sometimes you don’t know where to start or what to ask—especially on a topic like anomaly detection that you’re still becoming familiar with. In that case, it’s best to listen to the questions of others and let their line of thinking guide your learning. Below are some questions we received during our “Ask Me Anything: Anomaly Detection” webinar to help you get started.

What’s the difference between outliers and anomalies?

Outliers are observations that are distant from the mean or location of a distribution. However, they don’t necessarily represent abnormal behavior or behavior generated by a different process. On the other hand, anomalies are data patterns that are generated by different processes.

Are there any applications for anomaly detection in pharmaceuticals?

There are many applications of anomaly detection in the pharmaceutical life sciences space. Including, process monitoring and quality control using Statistical Process Control (SPC) or Quality Control (QC) and Multivariate SPC (MSPC) charts in pharmaceutical manufacturing. Timely anomaly detection is critical to avoid abnormal events and adhere to safety standards. Identifying anomalies in over-the-counter transactions can be used to fight prescription abuse in pharma retail data. And real-time detection of anomalies in multi-parameter clinical trial data helps ensure the success of clinical trials.

Are GANs also used for anomaly detection? If so, could you please provide an industry use case.

Generative Adversarial Networks (GANs) are novel unsupervised learning methods that are very efficient in identifying anomalies. Since GANs are iterative by design and the adversarial training is optimized to leverage reconstructed samples to reduce residual loss, they work well in semi-structured and unstructured data. They are specifically very useful for medical image analysis (helping radiologists find hard-to-identify tumors), facial recognition, text to image translation, etc.

Does correlation data affect anomaly detection? In what ways and how could we diminish the effects? Is it preferable to clean and remove correlations before starting anomaly detection?

As mentioned in the webinar, we do not think that correlations would affect anomaly detection, but we have many techniques available to help determine how to treat correlated variables. One suggestion would be to reduce the number of dimensions with a technique like Principle Component Analysis (PCA).

What is a suggested algorithm that would be appropriate for anomaly detection related to identifying unusual activities in network activities or data?

As mentioned in the webinar, there are many methods and algorithms that work well for various applications and use cases of anomaly detection. Some of them are Recurrent Neural Network (RNN), Generative Adversarial Network (GAN), Isolation forests, Deep Autoencoders, etc. If you are specifically interested in Network/Graph analytics, the two main methods used for identifying anomalies in network graphs are the Direct Neighbour Outlier Detection Algorithm (DNODA) and Community Neighbour Algorithm (CNA).

In my current work “novelty” is the main thing we struggle to detect. Quality control charts work well for known patterns, but automated identification of new patterns is harder. I’m hoping to get some ideas for tools to help in that area.

For univariate quality control charts, Western Electric rules are used to detect a handful of common patterns. Classic multivariate methods such as Partial Least Squares (PLS) will capture patterns involving more than one variable that would not be detected by univariate methods. Autoencoder is the most comprehensive tool that will cover the widest range of different patterns. It can capture patterns that are multivariate, cyclical, nonlinear, and have interactions. You train the autoencoder using a set of ‘normal’ data’; any patterns present in new data that were not present in the training set will be flagged. We have a template for Spotfire on the TIBCO Community.

Would reducing the dimensions by doing PCA affect the anomalies in a dataset? Would it lead to the disappearance of the anomalies? If so, how could it be prevented?

Doing PCA will capture some percentage of the variance in the original dataset. So the way we use PCA to do anomaly detection is to compute the ‘distance’ from the original point to the point represented in the lower-dimensional space. The larger the distance (i.e. the more that was ‘lost’ when mapping the observation to the lower-dimensional space), the more we regard it as an anomaly.

Chances are you had a similar question to the ones above, but didn’t know how to ask it. In fact, if any of the questions above spark another one in your mind, contact us and we can answer it for you. And watch the full on-demand “Ask Me Anything: Anomaly Detection and Machine Learning” webinar for more on the basics of anomaly detection, common use cases, and some key techniques to keep in mind as you get started.

Let’s block ads! (Why?)

The TIBCO Blog

Spares Pooling: Machine Learning To The Rescue

December 28, 2019 SAP

Spare parts. Every original equipment manufacturer (OEM) has to maintain an inventory of them. Machines break, pieces of equipment fail, and customers need spares to make repairs without delay.

In today’s on-demand global market, spares replenishment must work like magic: prompt, cost-effective, and seamless. Otherwise, prolonged downtime due to equipment failure can have serious direct and indirect impacts, for example, dysfunctional surgery suites in hospitals, flight cancellations, production delays, unfulfilled orders, or poor quality leading to product recalls.

Ensuring the right parts are on hand when needed

Having a sufficient and efficient supply of spare parts is a vexing inventory-management scenario. Forecasting parts-failures at the points of demand and planning for replenishment are essential, as parts are very expensive and costly to keep in stock. Meanwhile, warehouse management teams must routinely juggle multiple, complex scenarios with contradictory insights, forecast failures, and inventory based on complicated predictive analytics.

Demand and consumption forecasting is done at the stockkeeping unit (SKU) level by taking into account the way they handle failures in KPIs. Doing so can minimize the costs of downtime. The key is to apply a risk-ranking system estimates the consequences of a failure to the business, the probability that a failure will occur, and the risk the business is willing to accept.

In terms of a repair strategy, a lot depends on the OEM’s policy on parts replaceability versis repairability. Availability of onsite skills within the maintenance team and time to repair are crucial factors in these decisions. Sometimes, it is more cost-effective to replace the unit and to repair the old unit at one location. But stocking whole units is far more expensive than individual SKUs.

Calculating the efficient safety stock levels at the SKU level must take into account the predictability of demand, delivery lead times, delivery accuracy, obsolescence, transportation costs, and more. The strategy must include reorder points (ROP) and reorder quantities (ROQ), as well as cost-based optimization of order quantities and order cycles at the supplier or product level. Combining automated routine orders with order suggestions for specific important items as well as automatic exception management must also be considered.

Pooling dramatically changes the stocking strategy

Most OEMs have multiple warehouses with millions of dollars’ worth of inventory. If they were to stock at least one piece of each part in all these warehouses, the cost would be prohibitive. This is where parts pooling comes in.

Pooling builds on data that proves aggregation reduces variability and uncertainty. This implies that demand variability will decrease by aggregating spares across locations.

For example, low demand from one warehouse location offsets the demand at another warehouse location. This decline in variability allows a reduction in safety stock and reduces average inventory. When executed well, it is an excellent approach to reduce inventory at individual warehouses without impacting customers.

Shared and pooled warehousing have the potential to transform spares management of small and midsize businesses in competitive markets. This is quite valuable as business networking strategies and collaboration in logistics continue to gain momentum.

Easier said than done

Location-base parts pooling requires adding many highly complicated variables to the current set of variables. In this case, predictive modeling is no longer sufficient. It is time to move to artificial intelligence (AI) and machine learning models that take into account multiple variables and blend them to enable better decision-making in two critical areas:

Warehouse location and product flow: Decisions on whether to have many warehouses close to customers or to have more centralized locations can balance the advantages of aggregated demand and lead times. Proximity to customers and other factors may make maintaining more warehouses more advantageous.

Transportation: Managing parts pooling across locations means accounting for the geographical area along with logistics like availability and reliability of the ground transportation network, flights, or water corridors.

Within months of implementing warehouse pooling optimized through machine learning, most OEMs’ customer KPIs improve while simultaneously minimizing the capital invested in spares.

AI and machine learning models merge cost, operations, and criticality analysis for the relevant scenarios while factoring location criteria at a granular level to automate actions and enable better decisions. Optimization enable OEMs to reduce inventory without impacting customers by using a pooling model.

To reduce risk, OEMs can start by implementing an ML model with a few fixed locations and fewer optimization criteria, like the cost of parts, shipping costs, lead times, and the probability of failures. As the system matures, they can improve this model through a dynamic pooling model capable of handling more factors.

Lenient spares inventory management is a strict no-no as the costs of non-availability can be disastrous. Machine learning-enabled systems are set to transform the way warehouses operate. But before OEMs take the plunge, they must make sure they have the data and talent they need.

See some real-world examples of how machine learning is establishing itself as a mainstream technology tool for companies, enabling them to improve productivity, planning, and ultimately, profits, in the study “Machine Learning In The Real World.”

Let’s block ads! (Why?)

Digitalist Magazine

The Benefits of Doing Machine Learning and Analytics in the Cloud

December 17, 2019 TIBCO Spotfire

Reading Time: 4 mins 30 sec.

Analytics with data science has been one of the last enterprise systems to move to the cloud, but the situation has changed fundamentally in just the last year or two.

Suddenly, there is a proliferation of cloud-based databases and open-source machine learning development frameworks like SageMaker and TensorFlow—all of them now being heavily promoted by the major cloud vendors (Amazon, Microsoft, Google, and more).

The cloud is quickly becoming everyone’s preferred way of doing machine learning and analytics. If you know your way around all the available components, it can be easy to build even the most sophisticated machine learning models for everything from image recognition to fraud detection in the cloud.

What to use—when and how

There’s a ton of rich functionality available in the cloud that you can spin up right now. Over the last few years, there’s been a real shift from heavyweight on-premises installations of data science and predictive analytics to the more lightweight approach that the cloud offers.

The breadth of capabilities that the cloud providers have created combined with the ease of use of a data science platform like TIBCO’s, organizations can spin up environments very quickly without a great deal of IT overhead. This combination of scalability and flexibility is the central value of the cloud when it comes to doing analytics. In fact, with TIBCO^® Data Science, you can create solutions across all of these various cloud environments without needing to learn the nuances of each.

Here’s a helpful chart of the technologies available for artificial intelligence (AI) and machine learning in the cloud:

The Benefits of Doing Machine Learning and Analytics in the Cloud

Proof it’s as easy as it sounds

It really is that easy. The proof is in our customers’ success stories. Below are a couple of case studies that show how easy it can be to build applications based on sophisticated AI and machine learning, using the cloud.

Tipping Point Community fights poverty with data:

As a non-profit organization looking to better understand the drivers behind poverty, Tipping Point started a project to explore correlations between parking citations, late fees, and low-income individuals. Using TIBCO Data Science’s collaborative interface for business users and deploying machine learning models in the cloud to discover insights, Tipping Point found a disproportionate impact on low-income drivers. We’re proud to say these data-driven recommendations that Tipping Point made to the San Francisco Office of Financial Justice led to a change in policy to make the system fairer.

Leidos unlocks big data potential for healthcare analytics:

Leidos partnered with TIBCO Data Science, an enterprise-class cloud platform that leverages Amazon Web Services (AWS), to allow users to create machine learning workflows. By using the cloud, Leidos opened up collaboration across teams and was able to perform quicker analyses. It was able to analyze healthcare data to determine the cause of disease outbreaks like HIV and Zika, consolidate data around emerging healthcare policies, and explore human factors affecting space exploration for NASA.

Across these and many other examples, performing analytics and machine learning in the cloud gives organizations the ability to uncover hidden patterns, anticipate outcomes, and react quickly to real-world events. Data science teams can spin up new systems in the cloud in a matter of hours, performing advanced analytics in a low-code, visual data science environment like TIBCO’s to find the answers to their toughest problems, fast.

Connect teams & scale algorithms with cloud-enabled tech

The TIBCO Data Science platform provides an interactive interface for teams to collaborate on projects in the cloud. Teams scattered around the globe, across different departments, in different roles, can connect through the web-based interface to solve difficult data science problems together. Furthermore, the algorithms TIBCO provides through a simple drag and drop interface are not only easy to use, without requiring a lot of code, but they’re also readily scalable and immensely powerful. But if coding is your thing, you can also use the embedded Jupyter Notebooks that are built into the platform.

Watch this webinar for more on available cloud technologies and the many benefits of doing machine learning and analytics in the cloud.

Get started right now. Go to the AWS marketplace and sign up for a preview of the TIBCO Data Science platform, connect to open sources of data, and start building machine learning models today.

Let’s block ads! (Why?)

The TIBCO Blog

Tag Archives: Learning

Yann LeCun and Yoshua Bengio: Self-supervised learning is the key to human-level intelligence

UC Berkeley researchers open-source RAD to improve any reinforcement learning algorithm

Health care organizations use Nvidia’s Clara federated learning to improve mammogram analysis AI

Lendio raises $55 million to surface small business loans using machine learning

SQL Server Machine Learning 2019: Working with Security Changes

Launchpad in SQL Server 2017

Launchpad in SQL Server 2019

Proving this Concept

Solutions

All Application Containers

Using the App Containers SID

Workaround with a FileTable

Enabling Filetable in SQL Server

Preparing the Database to Support Filetable

New R Script for Filetable

Conclusion

Researchers propose data set to measure few-shot learning performance

AI Weekly: Machine learning could lead cybersecurity into uncharted territory

7 Questions on How to Use Machine Learning for Anomaly Detection

What’s the difference between outliers and anomalies?

Are there any applications for anomaly detection in pharmaceuticals?

Are GANs also used for anomaly detection? If so, could you please provide an industry use case.

Does correlation data affect anomaly detection? In what ways and how could we diminish the effects? Is it preferable to clean and remove correlations before starting anomaly detection?

What is a suggested algorithm that would be appropriate for anomaly detection related to identifying unusual activities in network activities or data?

In my current work “novelty” is the main thing we struggle to detect. Quality control charts work well for known patterns, but automated identification of new patterns is harder. I’m hoping to get some ideas for tools to help in that area.

Would reducing the dimensions by doing PCA affect the anomalies in a dataset? Would it lead to the disappearance of the anomalies? If so, how could it be prevented?

Spares Pooling: Machine Learning To The Rescue

Ensuring the right parts are on hand when needed

Pooling dramatically changes the stocking strategy

Easier said than done

Recent Posts

Categories

Archives