Machine Learning
1297 Threads
When you learn data science, thereβs no replacement for project-based learning. Courses and blogs are fantastic resources, and you should take advantage of them. But if you ever...
Tuned in PyTorch Turns 5 video! I have transcribed @ylecun's prediction about AI for the next five years. It's enlightening. See thread below π𧡠#PyTorch #julialang #ArtificialI...
Graph neural networks (GNNs) are rapidly advancing progress in ML for complex graph data applications. Let's have a look at some resources to help you learn and keep up-to-date wi...
What matters most when training a neural network is how well it generalizes to unseen data. For neural networks, it turns out there's a simple principle that can allow you to unde...
Wifey Research Thread 3 ****************************** 1. Simple Asset Allocation 2. Short-Term / Intraday 3. Machine Intelligence
Linear Regression in Machine Learning A Threadππ§΅
What are the most intriguing areas of deep learning? Let's use the examples of the biggest AI labs in the world. What do they believe is key to the near future of AI? 7 companies...
As amazing as it is, there are only five questions it can answer. Source: @_brohrer_ [https://t.co/lZzIz8RaGp] https://t.co/z65Rm81TlW
A lot of machine learning research has detached itself from solving real problems, and created their own "benchmark-islands". How does this happen? And why are researchers not esc...
Tired of there being 5+ types of data "shift" that interchange meaning in different ML research papers or Medium posts. E.g. covariate shift, domain shift, concept shift, subpopula...
RL agents can make decisions in 2 basic ways: 1) model-free 2) model-based Let's discuss model-based RL methods. - How do they work? - Categories of model-based RL - Advantages +...
GPT has been a core part of the unsupervised learning revolution thatβs been happening in NLP. In part 2 of the transformer series, weβll build GPT from the ground up. This thread...
Hyperparameters are numbers that the model cannot learn. Finding the right hyperparameters is crucial for the viability of your ML models. 2 Blackbox Hyperparameter Optimizations...
πΎ Machine Learning: Algorithms, Models, and Applications (154 pages) π π Book: https://t.co/Da94CR8Fqo π½ Abstract π½ https://t.co/oKKOPdIXcF
Forecasting high-dimensional time series plays a crucial role in many applications like: - demand forecasting - financial predictions You can use @AmazonScience's DeepGLO for the...
How do transformers work with long text sequences? They're pretty challenging for transformers. It is due to the self-attention mechanisms transformers use. Here is how @allen_ai...
[π§ Paper Summary π] An interesting paper was recently published to arxiv: "Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasets" (although it originally appear...
@parlai_parley is a unified framework for sharing, training, and testing dialog models. 5 tips: 1) 100+ popular datasets 2) A wide set of reference models 3) Vast model zoo 4) Use...
If you want to understand why TensorFlow is the way it is, you have to go back to the ancient times. In 2012, Google created a system called DistBelief that laid out their vision f...
Uncertainty estimation is one of the most difficult challenges. But it regularly affects the performance of time-series forecasting. @Uber Research proposes a method based on Bay...
Today Let us understand and implement the basic mechanism/flow of the ML algorithm. Here are the steps for one training loop. β£ Perform a prediction. β£ Compute error/loss. β£ Up...
Transformers are arguably the most impactful deep learning architecture from the last 5 yrs. In the next few threads, weβll cover multi-head attention, GPT and BERT, Vision Transf...
πΉ 7 #DataScience Projects You Should Do to Make Your Resume Stand Out π 1. Regression Project 2. Classification Project 3. Clustering Project 4. Sentiment Analysis Project 5. Rec...
Lots of great research papers in ML and NLP in 2021. Let's take a look at my favorite papers this year (one for each month) β