Machine Learning and Deep Learning Open Lab Series

Welcome to the Machine Learning and Deep Learning Open Lab Series! This series offers seven open labs to introduce participants to the core concepts and techniques in Machine Learning (ML) and Deep Learning (DL). These open labs will prioritize intuitive understanding of machine learning algorithms and deep learning approaches. These are intended to complement machine learning and deep learning courses taught at ASU, focusing on practical applications and intuitive explanations of difficult concepts and examples.

Open Lab 1: Introduction to Machine Learning

Description:

This open lab will cover an overview of fundamental concepts, applications, and career opportunities in ML. We will also explore how to get started with learning ML and explore the possibilities in exciting field of DL that produced all recent AI developments.

Key Topics:

• Introduction to Machine Learning
• Types of ML: Supervised, Unsupervised, and Reinforcement Learning
• Overview of popular ML models and their use cases
• Career paths in ML and how to start
• No coding

---

Open Lab 2: Supervised Learning - Classification Models

Description:

This open lab will explore classification models for predicting categorical outcomes. We will also review metrics to assess classifiers and improve the accuracy of classification algorithms.

Key Topics:

• Logistic Regression
• Decision Trees
• Naive Bayes
• Support Vector Machines (SVMs)
• KNN
• More Other classifiers
• Evaluation metrics: Confusion Matrix, Accuracy, Precision, Recall, F1-Score etc

---

Open Lab 3: Supervised Learning - Regression Models

Description:

This open lab will cover regression techniques for predicting continuous numerical outcomes. We will also go through model evaluation.

Key Topics:

• Linear Regression and Polynomial Regression
• Regularization techniques (Lasso, Ridge)
• Decision Trees
• Random Forest
• Evaluation metrics: Mean Squared Error (MSE), Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), R²

---

Open Lab 4: Unsupervised Learning and Dimensionality Reduction

Description:

This open lab will explore techniques for working with unlabeled data and reducing complexity of datasets. It review clustering and dimensionality reduction methods to uncover hidden patterns in data.

Key Topics:

• Clustering methods: k-Means and Hierarchical Clustering
• Dimentionalty Reduction: Principal Component Analysis (PCA)

---

Open Lab 5: Advanced ML - Ensemble Methods and Model Optimization

Description:

In this open lab, we will cover advanced ML techniques like ensemble methods and hyperparameter optimization to boost model performance.

Key Topics:

• Ensemble methods: Random Forests, Gradient Boosting (XGBoost, LightGBM)
• Hyperparameter tuning with GridSearchCV and RandomizedSearchCV
• Cross-validation and model evaluation

---

Open Lab 6: Deep Learning and Neural Networks

Description:

We will cover building blocks of deep learning and neural networks and explore backpropagation, optimization, and architecture design.

Key Topics:

• Neural Network architecture and activation functions
• Backpropagation and optimization techniques (SGD, Adam)
• Types of Neural Networks
• Classification example with Feedforward Neural Networks

---

Open Lab 7: Transformer-based Models

Description:

This open lab will introduce participants to the fundamentals of new AI developments, focusing on modern techniques. It will cover key topics such as transformer architecture, fine-tuning, and transfer learning, providing a strong foundation in new AI methods. We will also explore Large Language Models (LLMs), their future, and the resurgence of RNNs.

Key Topics:

• Transformer Architecture
• LLMs
• Pretrained models (e.g. BERT)
• Future of AI

---