Machine Learning Basics: Python, Numpy & Scikit-Learn

Course Overview

• Course Title: Machine Learning Basics: Python, Numpy & Scikit-Learn
• Instructor: Social Science Academy
• Target Audience:
  • Beginners in machine learning
  • Individuals with a basic understanding of Python programming
  • Those curious about machine learning applications
  • People interested in molecular data analysis
• Prerequisites:
  • No prior experience in machine learning or data science is required
  • Familiarity with using Google Colab or Jupyter Notebooks is helpful (not mandatory)
  • Basic understanding of Python programming (variables, loops, functions)
  • A working laptop or desktop with internet access

Curriculum Highlights

• Key Topics Covered:
  • Core ML concepts: loss functions, gradient descent, epochs, learning rates
  • Linear and logistic regression models
  • Numpy, Pandas, Matplotlib, and Scikit-learn
  • Neural networks implementation
  • Handwritten digit classification using MNIST dataset and Keras
  • Regularization techniques: early stopping, dropout
  • Molecular data analysis using RDKit
  • Graph convolution techniques using MolGraph
  • DeepChem for molecular dataset training
• Key Skills Learned:
  • Building and training linear and logistic regression models from scratch in Python
  • Working with real datasets using Numpy, Pandas, Matplotlib, and Scikit-learn
  • Implementing neural networks step-by-step, including forward and backpropagation
  • Classifying handwritten digits using MNIST dataset and Keras
  • Preventing overfitting with regularization techniques
  • Analyzing molecular data using RDKit and visualizing chemical structures
  • Applying graph convolution techniques to molecular structures using MolGraph
  • Training models on molecular datasets using DeepChem

Course Format

• Duration: 7 hours on-demand video
• Format: Self-paced online course
• Resources:
  • Access on mobile and TV
  • Certificate of completion