Khetha0121/AnalystLabs-Africa-Machine-Learning

# AnalystLab Africa Machine Learning Internship Program ### Batch B | June – August 2026 --- ## Overview This repository contains my work for the **AnalystLab Africa Machine Learning Internship Program (Batch B)**. Each week covers a different stage of the machine learning pipeline — from data preprocessing and EDA through to model development and evaluation. --- ## Repository Structure ``` AnalystLabs-Africa-Machine-Learning/ │ ├── AnalystLab_EDA_Assignment.ipynb # Full Jupyter Notebook (executed) ├── AnalystLab_EDA_Report.md # Summary report in Markdown ├── titanic_cleaned.csv # Cleaned Titanic dataset ├── iris_cleaned.csv # Cleaned Iris dataset │ └── outputs/ ├── titanic_bivariate.png # Titanic bivariate analysis chart ├── titanic_corr.png # Titanic correlation heatmap ├── iris_bivariate.png # Iris bivariate analysis chart ├── iris_corr.png # Iris correlation heatmap ├── iris_pairplot.png # Iris pairplot └── iris_cleaned.csv ``` --- ## Weekly Progress ### Week 1–2: Data Preprocessing & Exploratory Data Analysis (EDA) **Datasets used:** - Titanic Dataset — Binary classification (Survived / Not Survived) - Iris Dataset — Multiclass classification (Setosa / Versicolor / Virginica) **Work completed:** - Loaded and inspected both datasets using Pandas - Identified and handled missing values (median/mode imputation) - Detected and treated outliers using boxplots and percentile capping - Encoded categorical variables (Label Encoding & One-Hot Encoding) - Applied feature scaling (StandardScaler & MinMaxScaler) - Performed univariate and bivariate analysis with visualisations - Generated correlation heatmaps and pairplots - Summarised ML readiness for both datasets **Key Insights:** | Dataset | Top Finding | |---------|------------| | Titanic | Sex (gender) was the strongest predictor of survival (r = 0.54) | | Iris | Petal Width was the most discrimin …