This is first part of 34 part series of notes on deep learning.

This is first part of 92 part series of conventional guide to supervised learning with scikit-learn written with a motive to become skillful at implementing algorithms to productive use and being able to explain the algorithmic logic underlying it. Supervised learning In supervised learning, each example is a pair consisting of an…

The complete guide to NLP with fastai — Topics Covered: 1. What is NLP? (click the link to the article ) Topics covered: A changing field, resources, tools, python libraries, example applications, ethics issues 2. Topic Modeling with NMF and SVD part-1: Topic Modeling with NMF and SVD: part 1 part-2 : Topic Modeling with NMF and SVD: part 2 (click the link to the article) Topics…

The Theory of Portfolio Management blog series presents my favorite selected list of research articles related to essential portfolio management topics like factor analysis, theory of asset allocation, risk management, and position sizing. Theory of Portfolio Management Part 0: Prologue and links to articles in this series

Anchors are high precision explainers that use reinforcement learning methods to come up with the set of feature conditions (called anchors), which will help explain the observation of interest and also a set of surrounding observations with a high precision (the user is free to choose their minimum precision cut-off). — Anchors Anchors is a model explanation technique that provides understandable and human-interpretable explanations for machine learning models’ predictions. Anchors work by generating “if-then” rules that explain why a model made a particular prediction. These rules are simple, and interpretable, and provide transparency to the decision-making process of the model.

LRP is a technique that brings such explainability and scales to potentially highly complex deep neural networks. It operates by propagating the prediction backward in the neural network, using a set of purposely designed propagation rules — Layer-wise Relevance Propagation (LRP) Layer-wise Relevance Propagation (LRP) is a model explanation technique that aims to provide insights into the decision-making process of neural networks. LRP works by propagating the final output of a neural network back through its layers to determine the relevance of each input feature to the output.

LIME is a visualization technique that helps explain individual predictions. It is model agnostic so it can be applied to any supervised regression or classification model. — Local Interpretable Model-agnostic (LIME) Local Interpretable Model-agnostic (LIME) is a model explanation technique that provides insight into the decision-making process of complex machine learning models, without requiring any prior knowledge of the model’s architecture or parameters. LIME works by generating interpretable models that are locally faithful to the predictions of the original model.

Algorithm that uses the gradients of any target concept (say ‘dog’ in a classification network or a sequence of words in captioning network) flowing into the final convolutional layer, to produce a coarse localization map highlighting the important regions in the image for predicting the concept. It works only with CNNs. — Gradient-weighted Class Activation Mapping (Grad-CAM) Gradient-weighted Class Activation Mapping (Grad-CAM) is a visualization technique for deep neural networks that helps to understand which parts of an image were important for the model’s decision-making process. It works by generating a heatmap that highlights the regions of an image that were most influential in the model’s output.

Explainable AI (XAI) aims to create AI systems that can provide clear explanations for their decisions and actions. This is important for ensuring transparency and accountability in AI systems, as well as building trust with users. — In this blog, we will see several techniques and approaches that can be used to achieve explainability in AI systems and then see four XAI algorithms in separate four blogs linked below