Causal Inference on Loyalty Program Effect on Churn

🚀 Day 04 of My Machine Learning Journey: NumPy Data Types (dtypes) Today, I learned about NumPy data types (dtypes), which define the type of elements stored in an array. I explored: ✅ Different types like int, float, and bool ✅ How NumPy uses fixed data types for better performance ✅ Why choosing the right dtype helps optimize memory usage Understanding dtypes helps write more efficient and faster code — an important step for Machine Learning. 💡 #MachineLearning #NumPy #Python #LearningJourney #Day04

📊 Day 6 | K-Nearest Neighbors (KNN) 🤝📍 Today, I learned about K-Nearest Neighbors (KNN), a simple and intuitive Machine Learning algorithm. KNN works on the idea of distance — it classifies a data point based on the majority class of its nearest neighbors. 📌 In simple terms: “Similar data points are close to each other.” Example: ✔ Recommending products ✔ Classifying customers To understand this, I implemented KNN using Python and observed how it predicts based on nearby data points 💻 KNN is simple but powerful for many classification problems. #MachineLearning #KNN #DataScience #LearningInPublic #Python

Just wrapped an energizing session teaching data loaders for local LLMs in Python! We mapped out Text Loaders pulling PDFs, JSONs, CSVs, and TXTs into strings or key-value trees via #LangChain, Image Loaders converting to binary formats, plus #OCR magic on pure-image #PPTX files using Python’s pptx loader—and Pandas crushing CSV/XLS flows. Total game-changer for building rock-solid AI pipelines! Tell me how do you think about texts on whiteboard? #Python #AI #LocalLLM #whiteboardKnowledge

Day 109 Backtracking patterns are repeating again — and that’s a good sign. #Day109 🧩 78. Subsets How today went: • Used recursion to explore all elements • At each step, decide to include or skip the current element • Append current subset → explore → then pop to backtrack • Move to the next index and repeat What I’m noticing: Subsets is one of the cleanest backtracking patterns: → choose → explore → undo Another revision day, but clarity is improving. Consistency continues. #LeetCode #DSA #Python #Backtracking #Recursion #LearningInPublic #Consistency

Many people learn Python and Pandas as tools. But the real transformation happens when you learn Pandas as a way of thinking. Because data isn’t just “numbers in a table”—it’s evidence. And evidence has shape, structure, friction, and sometimes silence (missing values, messy formats, inconsistent categories). When you master core Pandas operations, you stop merely processing datasets… and you start understanding systems. #Python #Pandas #LakkiData #LearningSteps

In my latest video, I break down the math behind logistic regression, derive the gradient descent update rules, explore vectorized implementations, and finally, code it from scratch in Python. Perfect for anyone preparing for ML interviews or looking to strengthen their foundations in machine learning. Video Link: youtu.be/cT_U40djaww   Channel Link: youtube.com/@datatrek

🔁 Exploring Sorting Algorithms in Python Today I practiced two fundamental sorting techniques: ✅ Bubble Sort ✅ Selection Sort 💡 Key Learnings: * Bubble Sort repeatedly swaps adjacent elements to push larger elements to the end * Selection Sort selects the minimum element and places it in the correct position * Understanding time complexity becomes clearer when you count operations manually #Python #DataStructures #Algorithms #CodingJourney #100DaysOfCode #LearningInPublic

Day7 of #30DayChartChallenge Theme: Multiscale Category: Distributions Tool: Python Data Source: python scikit-learn Datasets I worked with a few features from a machine learning dataset and plotted their distributions. At first, everything sits on different ranges. One stretches far, another stays tight, another somewhere in between. It looks fine, but comparing them like that is off. After scaling, they fall into the same range. Now the comparison actually makes sense. It’s a small step in most workflows, but seeing it visually makes the difference clearer. #30DayChartChallenge #python #Dataviz #Datascience

Subsets: Classic Backtracking Template Generate all 2^n subsets via binary decision tree — include or exclude each element. Base case: index exceeds array length, save current subset copy. Backtracking: add element, recurse, remove element (backtrack), recurse again. Critical Detail: subset.copy() is essential — without it, all results reference same list, causing incorrect final output. Each subset snapshot must be independent. Time: O(2^n) | Space: O(n) recursion #Backtracking #Subsets #DecisionTree #DeepCopy #Recursion #Python #AlgorithmDesign #SoftwareEngineering

🎬 Building my first Movie Recommendation System. Explored a dataset with 1.4M+ rows and found heavy missing values in key features like genres, keywords, and overview. My approach: dropping missing data to keep recommendations meaningful. What would you do? Drop or fill? Let me know your thoughts in the comments 👇 #MachineLearning #Python #DataScience