Welcome! This repository contains resources and materials that I used for learning and practicing machine learning concepts.
The Theory folder includes all the theoretical content covered in the course. This encompasses lecture notes, presentations, and any additional resources that provide a solid foundation in machine learning concepts.
The Lab folder consists of various practical exercises where I gained hands-on experience with different machine learning algorithms. Each lab comes with its own set of instructions and datasets, which allowed me to apply the concepts learned in the theory section.This folder helped me to sharpen my skills and develop a practical understanding of machine learning techniques.
The Project folder is the highlight of this course, offering a more in-depth and practical application of machine learning concepts. Inside this folder, you'll find two distinct projects:
- Description: Application of clustering techniques to perform customer segmentation for marketing purposes.
- Goal: Analyze and understand the main characteristics of a customer base to support marketing decisions.
- Problem Understanding: Join a data science and AI engineering team of a telco company to support decision-making in marketing and customer care.
- Scenario: Marketing colleagues plan to launch a new commercial campaign for a new mobile tariff and need insights into customer patterns.
- Algorithms Implemented:
- K-Means: Training the model and performance evaluation
- K-Means: Insights generation
- Mixture of Gaussians: Training and insights generation
- Clustering
- Description: Application of different supervised classification techniques in marketing and sales.
- Use Cases:
- Identify new customers in the market
- Identify customers in our internal Data Warehouse more likely to buy a new product
- Identify unsatisfied customers and potential churners
- Classify text into categories for spam identification or processing customer messages/emails.
- Machine Learning Process:
- Data understanding and preparation: Exploration of the dataset and feature engineering (missing values, outlier identification, categorical variables management)
- Model Training: Training baseline SVM and Decision Trees, analysis of metrics (recall, precision, confusion metrics), and improvement of classification through techniques like undersampling or ensemble of models.
- Creating a Business Opportunity with Machine Learning: Selection of the best model and identification of the most important features.