Account Share

Discover and read the best of Twitter Threads about #machinelearning

Most recents (15)
A thread for the "#software is not political" crowd. Whether you like it or not, all software is political because technology AFFECTS PEOPLE. If you came to #tech thinking you'd escape having to think about people, #politics, and society - you were mistaken. I'll show you why...
Before we start, remember this as you read. Just because YOU think the answer to any of these issues is clear-cut, it doesn't mean that issue is not political.

There are people who take the OPPOSITE position with just as much conviction, and they think it's clear-cut too.
Also - as an engineer, a developer or designer - if you choose to ignore the political and societal implications of your #technology, YOU HAVE MADE A POLITICAL CHOICE.
Read 12 tweets
#Thread on new features and updates announced in release notes of #MSDyn365 Oct'18:…. This is an excellent longread. Thread is just #TLDR for #CE specific/related features

#PowerApps #CDS #Flow #USD #Sales #FieldService #USD #Portals #AzureML
Starting with #MSDyn365 #Sales updates:
1. Playbooks: Think of it as Barney Stinson's playbook. Basically set of "automated repeatable sales activities" that help in winning opportunities. Looks like it'll be a set of activities which could be assigned to users
2. #MSDyn365 #Sales will feature deeper integration with LinkedIn, including capability to send InMail and adding LinkedIn related step in Business Process Flows

3. Out of the box @MicrosoftTeams integration
Read 31 tweets
OK people. I'm live tweeting @neo4j's #Chicago #GraphTour event today.
We are currently in the keynote. The speaker is talking about how popular Neo4j is
Keynote now telling us about all the new Neo4j features. Examples: location filter, including 3d. Auto cache reheating. I'm interested to know if auto cache reheating is working with query patterns or meant to replace them or what
Read 64 tweets
Did Thread Reader help you today?
Support me: I'm a solo developer! Read more about the story
Become a 💎 Premium member ($30.00/year) and get exclusive features!
Too expensive?
Make a small donation instead. Buy me a coffee ($5) or help for the server cost ($10):
Donate with 😘 Paypal or  Become a Patron 😍 on
Bayes’ Theorem Definitions:
The vertical bar | stands for "given that".
P = Probability.
A & B are events.
P(A) & P(B) are the probabilities of events A and B. Each event is separate from the other.
P(A|B) is the probability of A being true given that event B is true.
#SoDS18 #ML
Say we have 2 coolers at an owambe: Cooler A is filled with 10packs of small chops only. Cooler B has 5packs of small chops and 5packs of Asun. You are then asked to close your eyes and pick a pack out of one cooler, which pack would you pick? #MachineLearning #SoDS18
Because you know that we have more of small chops in both coolers, your brain is most likely going to tell you have picked a pack of small chops - even when your eyes are closed. This is not wrong.
#MachineLearning #SoDS18
Read 14 tweets
Wer das Thema Künstliche Intelligenz #KI und #MachineLearnning im Unterricht behandeln möchte, findet in diesem Thread ein paar Beispiele, interaktive Spiele und Experimente 🤓🤖👇
#digitaleBildung #schule #twitterlehrerzimmer #unterricht
Künstliche Intelligenz wird zum Beispiel bei der automatischen #Bilderkennung eingesetzte. Auf kann man Gegenstände in die Kamera halten, ein Foto machen und die #KI interpretiert die Bilddaten 📸 🤓
Auch unter können Bilder hochgeladen werden - hier werden sogar noch mehr Informationen angezeigt und einzelne Objekte und Gesichter in den Bildern erkannt (Bildanalyse von Google) 👇🤖 #KI
Read 21 tweets
@francesc @vadimlearning First #MLonCode paper presented by @vadimlearning is "code2vec: Learning Distributed Representations of Code" by Uri Alon, Meital Zilberstein, @omerlevy_, and @yahave.
@francesc @vadimlearning @omerlevy_ @yahave The second #MLonCode paper, also presented by @vadimlearning, is "A General Path-Based Representation for Predicting Program Properties" by ... wait for it ... Uri Alon, Meital Zilberstein, @omerlevy_, and @yahave 🎉
@francesc @vadimlearning @omerlevy_ @yahave Next, our #ML intern Romain Keramitas, is reviewing the #MLonCode paper "Mining Idioms from Source Code" by Miltiadis Allamanis and @RandomlyWalking
Read 15 tweets
How many random seeds are needed to compare #DeepRL algorithms?

Our new tutorial to address this key issue of #reproducibility in #reinforcementlearning




#machinelearning #neuralnetworks
Algo1 and Algo2 are two famous #DeepRL algorithms, here tested
on the Half-Cheetah #opengym benchmark.

Many papers in the litterature compare using 4-5 random seeds,
like on this graph which suggests that Algo1 is best.

Is this really the case?
However, more robust statistical tests show there are no differences.

For a very good reason: Algo1 and Algo2 are both the same @openAI baseline
implementation of DDPG, same parameters!

This is what is called a "Type I error" in statistics.
Read 11 tweets
Today we're at #FOSD, talking about the future of software development with influential individuals in the fields of #MLonCode, #QuantumComputing, and #blockchain technologies.

Follow this thread for live tweeting!
Amazing talk by @KentBeck on how hundreds of thousands of developers could collaborate.

Moving from text transformations to tree transformations. Let's move away from "lines", which come from punch cards, and evolve into something that scales better.

Read 20 tweets
What #machinelearning models are used at hedge funds and other investment firms? Audience (200 ppl) poll at @Bloomberg ML in Finance event hosted by @jbaksht
And in what areas...
And in what asset classes.
Read 3 tweets
More hype around AI, mHealth, wearables. @AliveCor devices can use #AI to find LQTS:…
From the press release:

AUC 0.83, specificity 81%, sensitivity 73%, accuracy 79%

Prevalence of LQTS 160,000 out of US population of 325.7 million (0.049%)
Do we need to do a tweetatorial about how this is grossly insufficient accuracy and will almost certainly lead to more harm than good?
Read 51 tweets
(1/5) I grew up in the Programming Languages research community and have recently begun attending Machine Learning conferences. One perspective that I don't see much in either community is that #MachineLearning is a form of #programming.
(2/5) PL/formal methods researchers tend to think of programs as engineered objects, and study abstractions/tools for principled engineering. But the big assumption here is that you can formalize your goals and the world in which your programs run. That's not always realistic.
(3/5) In contrast, #MachineLearning lets programs be "found" objects. "I don't have a full spec for my program and can't write the code myself, but here's some data on what it does. Discover it!" This is still #programming, albeit done inductively rather than deductively.
Read 5 tweets
Read 25 tweets
1/6 Data and the Quest for Truth: A short cartoon talking about the value of data in establishing truth. First, distinguish between matters of opinion and matters of fact.

#datascience #scientificmethod #machinelearning

#think2018 #IBMPartner @ibmanalytics
2/6 You can have a preference and that’s cool. You like one thing. I like another. We run into problems when we argue about verifiable facts. Watch out for #confirmationbias

#datascience #analytics #confirmationbias

#think2018 #IBMPartner
3/6 Ignoring evidence can impact people around you… and that’s bad. We live in a #BigData in which everything is quantified. To solve our problems, we need to become data-literate.

#datascience #machinelearning

#think2018 #IBMPartner
Read 6 tweets
Inspired by the big ol' long list of deep learning models I saw this morning, and @SpaceWhaleRider's love of science-y A-Z lists, I've decided to create an A to Z series of tweets on popular #MachineLearning and #DeepLearning methods / algorithms.

Ready? Here we go:
A is for... the Apriori Algorithm!

Intended to mine frequent itemsets for Boolean association rules (like market basket analysis). Ex: if someone purchases the same products as you, in general, then you'd probably purchase something they've purchased.…
B is for... Bootstrapped Aggregation (Bagging)!

This is an ensemble meta-algorithm designed to improve the stability and accuracy of machine learning algorithms used in statistical classification+regression. Reduces variance, helps to avoid overfitting.

Example: Random Forests.
Read 28 tweets
1/ A new episode of #Perspectives! How will automation impact the industry, and you personally, over the next twelve months?…
2/ How will automation impact the industry and @dbrowell?… #mrx
3/ How will automation impact the industry and @andybuckers?… #newmr
Read 14 tweets
Related hashtags
Did Thread Reader help you today?
Support me: I'm a solo developer! Read more about the story
Become a 💎 Premium member ($30.00/year) and get exclusive features!
Too expensive?
Make a small donation instead. Buy me a coffee ($5) or help for the server cost ($10):
Donate with 😘 Paypal or  Become a Patron 😍 on