Primary Supervised Learning Algorithms Used in Machine Learning by Kevin Vu via @kdnuggets

In this tutorial, they are going to list some of the most common algorithms that are used in supervised learning along with a practical tutorial on such algorithms.

This is really useful and worth a bookmark or printout.

An Easy Guide to Choose the Right Machine Learning Algorithm by Yogita Kinha via @kdnuggets

There's no free lunch in machine learning. So, determining which algorithm to use depends on many factors from the type of problem at hand to the type of output you are looking for. This guide offers several considerations to review when exploring the right ML approach for your dataset.

This is really useful and in many ways, I wish it had been available years ago. Worth a bookmark.

Decision Tree Algorithm, Explained by Nagesh Singh Chauhan via @kdnuggets

All you need to know about decision trees and how to build and optimize decision tree classifiers.

A very clear and easy to understand guide that you might want to share with any folks that need the detailed information in it.

All Machine Learning Algorithms You Should Know in 2022 by Terence Shin via @TDataScience

Intuitive explanations of the most popular machine learning models.

This is really useful and definitely worth a read in case there is something new you haven't seen or come across yet. I particularly like that they are grouped into the type of algorithm.

New Algorithm Breaks Speed Limit for Solving Linear Equations by @KSHartnett via @QuantaMagazine

By harnessing randomness, a new algorithm achieves a fundamentally novel - and faster - way of performing one of the most basic computations in math and computer science.

I found this fascinating and it shows that progress is made all of the time - the tech world does not standstill.

K-Means 8x faster, 27x lower error than Scikit-learn in 25 lines by Jakub Adamczyk via @kdnuggets

K-means clustering is a powerful algorithm for similarity searches, and Facebook AI Research's faiss library is turning out to be a speed champion. With only a handful of lines of code shared in this demonstration, faiss outperforms the implementation in scikit-learn in speed and accuracy.

Definitely, a new one to try and see if you like the results better.

Algorithms For Data Scientists — Insertion Sort by Richmond Alake via @TDataScience

 One of the easiest algorithms you’ll ever learn. You might find the Wikipedia page on Insertion Sort too.

A good reminder of this algorithm and how to use it. Please note that this is only good for small datasets as it can be slow to build the final result one at a time.  Consider using Quicksort instead.

All Machine Learning Algorithms You Should Know in 2021 by Terence Shin via @TDataScience

Many machine learning algorithms exist that range from simple to complex in their approach, and together provide a powerful library of tools for analyzing and predicting patterns from data. If you are learning for the first time or reviewing techniques, then these intuitive explanations of the most popular machine learning models will help you kick off the new year with confidence.

This will help you get your machine learning right by using the correct algorithm.

Guide to Interpretable Machine Learning by @MatthewPStewart via @TDataScience

Techniques to dispel the black box myth of deep learning.

This is great and very detailed so put aside some time to read it as well as giving applause on the article.

AI Can Do Great Things—if It Doesn't Burn the Planet by/via @wired

OpenAI created an algorithm that successfully manipulates the pieces of a Rubik’s Cube using a robotic hand. But this accomplishment cost more than research time and effort—one estimate says it may have consumed about 2.8 gigawatt-hours of electricity, roughly equal to the output of three nuclear power plants for an hour. The computing power required for AI breakthroughs increased 300,000-fold from 2012 to 2018, creating an environmental impact that needs to be considered.

Something that we often don't consider but really should if we are serious about saving the planet.

5 trends to expect in the new big data protection revolution by Andrea Little Limbago via @infomgmt

Instead of regurgitating many of the dominant predictions around tech buzzwords such as quantum computing, 5G, IoT, the cloud, and artificial intelligence, let’s instead focus on the inherent duality of technology.

A great article to use to compare with your own plans and strategy.

WEBINAR: Train & Tune Your Computer Vision Models at Scale - 5 December 2019

Data Science Central Webinar Series Event

Train & Tune Your Computer Vision Models at Scale Join us for this latest DSC Webinar on December 5th, 2019 Whether you are training a self-driving car, detecting animals with drones, or identifying car damage for insurance claims, the steps needed to effectively train a computer vision model at scale remain the same. In this latest Data Science Central webinar, we’ll walk through best practices for managing a computer vision project including staffing, budgeting, and roles and responsibilities. Learn how to collect and label the data that will train and tune your machine learning algorithm, and which types of data labeling best fit your project along with the tools that will get the job done. In this webinar, you’ll learn how to: Identify key success factors when scoping a computer vision project Determine what kind of source data you need to make it successful Select tools that best fit your project Label your dataset so your algorithms can learn and perform as designed Speaker: Meeta Dash, Director of Product -- Figure Eight Hosted by: Stephanie Glen, Editorial Director -- Data Science Central Title:  Train & Tune Your Computer Vision Models at Scale Date:  Thursday, December 5th, 2019 Time:  9:00 AM - 10:00 AM PST Space is limited so please register early: Reserve your Webinar seat now

What a little more computing power can do for Deep Learning by Kim Martineau via @MIT

A deep learning model may need to see millions of photos before it can successfully identify a cat. The process is computationally intensive. But there may be a more efficient way - new MIT research shows that models only a fraction of the size are necessary.

An interesting viewpoint which could help to save money and time when developing this kind of model.

Facebook, Carnegie Mellon build first AI that beats pros in 6-player poker by Noam Brown via @facebookai

AI has never been good at bluffing, but Facebook’s Pluribus poker bot crushed human pros at six-player, no-limit Texas Hold’em. If each chip in the experiment were worth a dollar, Pluribus would have made $1,000 an hour against the pros. This is significant for real-world applications because unlike chess, poker is a hidden- or imperfect-information game—as are most real-world problems.

This was fascinating to me and well worth a read - you can also find some interesting stuff if you look on their website and could apply some of the techniques or disciplines to your own work.

The 5 Sampling Algorithms every Data Scientist need to know by @MLWhiz via @Medium

Here’s an intro to common sampling techniques.

Includes some sample Python code which makes it really easy to incorporate into your code (with some editing). Make sure you follow Rahul and give him lots of applause for helping you with this.

Open-endedness: The last grand challenge you’ve never heard of by Kenneth O. Stanley Joel Lehman and Lisa Soros via @OReillyMedia

While open-endedness could be a force for discovering intelligence, it could also be a component of AI itself.

This is a little bit of a long read but is worth the investment in time. A very interesting concept that I found fascinating. Something to think about.

All hail the algorithm by @Hey_AliRae via @AJEnglish

Al-Jazeera has published a five-part video series exploring the impact of algorithms on our everyday lives.

An interesting series and not something I would have associated with this channel.

The AI technique that could imbue machines with the ability to reason by Karen Hao via @techreview

“At six months old, a baby won’t bat an eye if a toy truck drives off a platform and seems to hover in the air. But perform the same experiment a mere two to three months later, and she will instantly recognize that something is wrong. She has already learned the concept of gravity.” Yann LeCun, the chief AI scientist at Facebook, hypothesizes that a lot of what babies learn about the world is through observation. And that theory could have important implications for researchers hoping to advance the boundaries of AI.

I definitely agree with his observation on the number of pictures needed for learning to generally take place which makes it NOTHING like the way a baby or young child would learn things in real-life. So unsupervised learning it is then.

Small example of k-means in R:

km <- kmeans(iris[,1:4], 3)
plot(iris[,1], iris[,2], col=km$cluster)
points(km$centers[,c(1,2)], col=1:3, pch=8, cex=2)
table(km$cluster, iris$Species)

How Etsy taught style to an algorithm by/via @FastCompany

Is it romantic or rustic? Boho or minimal? Etsy needed to offer searchers a way to find goods that matched their style aesthetics, but since descriptions aren’t uniform and don’t always describe the style, text mining the descriptions wasn’t enough. Colour and patterns don’t reliably predict style, so image recognition alone didn’t do it either. Enter a model that blends text analysis with image recognition based on 43 human-identified styles.

I love this real-life example detailing the steps they took to work out how to do this. Definitely, a methodology that could be used by other organisations to do a similar type of thing.

Where We See Shapes, AI Sees Textures by Jordana Cepelewicz via @QuantaMagazine

Deep learning vision algorithms often fail at classifying images because they take cues from textures, not shapes. This is a really interesting look at how machine vision actually processes the world.

This is absolutely fascinating and a great approach as to how relatively minor changes might make all the difference to your algorithms and outcomes.