Uma Mahesh

Naive Bayes is a family of probabilistic classifiers based on Bayes’ theorem and a strong conditional independence assumption among features. Despite the simplicity of that assumption, Naive Bayes remains one of the most effective, computationally efficient, and interpretable baseline classifiers…

K-Nearest Neighbors (KNN) is one of the most intuitive non-parametric supervised learning algorithms. It is used for both classification and regression, and it operates on a simple idea: similar observations tend to have similar outputs. Unlike models that explicitly learn…

Support Vector Machines are among the most elegant and mathematically grounded supervised learning algorithms. They are used for both classification and regression, and are especially valued for their geometric intuition, strong theoretical underpinnings, effectiveness in high-dimensional spaces, and the kernel…

Decision Trees and Random Forests are foundational tree-based learning methods used for both prediction and classification on structured data. They are especially valued because they can model nonlinear relationships, automatically capture interactions, work with mixed feature types, and offer a…

A detailed technical treatment of regression models for prediction and classification tasks, with HTML-formatted formulas suitable for WordPress. Abstract Linear Regression and Logistic Regression are two of the most foundational supervised learning algorithms in statistics, machine learning, econometrics, and predictive…

Techniques for Analyzing and Forecasting Sequential Data Abstract Time series data represents observations recorded sequentially over time. Unlike traditional datasets where observations are independent, time series data contains temporal dependencies, meaning past values influence future values. Such data is common…

Understanding GDPR and CCPA Compliance for Handling Sensitive Data Abstract Artificial Intelligence systems increasingly rely on large datasets containing personal and sensitive information. As organizations collect, process, and analyze data to build predictive models, concerns about privacy, security, and ethical…

Distributed Data Processing Frameworks for Large-Scale ML Abstract Modern machine learning systems often require processing massive datasets that exceed the capabilities of traditional data processing tools. As organizations generate vast amounts of structured, semi-structured, and unstructured data from digital platforms,…

Visualizing and Summarizing Data to Uncover Patterns, Relationships, and Anomalies Abstract Exploratory Data Analysis (EDA) is a fundamental step in the machine learning workflow that involves examining datasets to understand their structure, characteristics, and underlying patterns before building predictive models.…

Imputation Methods and Strategies for Handling Incomplete Datasets Abstract Real-world datasets are rarely complete. Missing values occur frequently due to errors in data collection, sensor failures, incomplete surveys, system migrations, or data corruption. If not handled properly, missing data can…