Roshmita Dey·Jul 29My Journey to Becoming a Research Data Scientist at GoogleFor the longest time, getting into Google felt like one of those goals that sits somewhere between “dream” and “statistically unlikely.” I…A response icon2A response icon2
Roshmita Dey·Apr 13Statistical Simulation in Python: A Practical Guide for Data ScientistsStatistical simulation is one of the most powerful tools in a data scientist’s toolkit. Whether you’re estimating probabilities, validating…A response icon4A response icon4
Roshmita Dey·Jan 24The 9 Most Critical Plots Every Data Scientist Must MasterIf you strip data science down to its core, it’s really about making decisions under uncertainty. Plots are the fastest way to reduce that…A response icon1A response icon1
Roshmita Dey·Jan 18Post-Training Validation: Cross-Validation vs BootstrappingHow to evaluate models reliably when data is limited, noisy, or expensive
Roshmita Dey·Jan 15Spectral ClusteringClustering is often introduced through algorithms like K-Means or Hierarchical Clustering. While powerful, these methods struggle when data…A response icon1A response icon1
Roshmita Dey·Dec 7, 2025Autoencoders for Anomaly DetectionAn autoencoder is a type of neural network designed to learn an identity mapping under a compression constraint.
Roshmita Dey·Dec 6, 2025What Is Product Analytics?In today’s data-driven world, building a product without analytics is like driving with your eyes closed. You might move fast — but you…
Roshmita Dey·Nov 28, 2025Regularized Linear Models Explained with Geometric IntuitionLinear Regression is often the first model we learn in Machine Learning. It’s simple, fast, interpretable — and dangerously easy to…
Roshmita Dey·Nov 27, 2025What Most People Forget About Regression AlgorithmsRegression is one of the first topics we learn in machine learning. Linear regression. Logistic regression. Ridge. Lasso. ElasticNet.
Roshmita Dey·Nov 24, 2025A Simple Guide to Choosing the Right ML Model Every TimeChoosing the right machine learning model can feel overwhelming — even for experienced data scientists. With dozens of algorithms and so…A response icon2A response icon2
