Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/zen204/-my-ecornell-portfolio

Includes all of my Jupyter Notebook assignments from MIT's Break Through AI/ML Machine Learning Foundations Program.
https://github.com/zen204/-my-ecornell-portfolio

area-under-curve dataframe feature-selection feature-selection-methods grid-search hyperparameters logistic-regression model-persistency optimal-value precision-recall receiver-operating-characteristic scikitlearn-machine-learning test-dataset training-dataset

Last synced: 29 days ago
JSON representation

Includes all of my Jupyter Notebook assignments from MIT's Break Through AI/ML Machine Learning Foundations Program.

Awesome Lists containing this project

README 

        
Unit 5: Evaluate and Deploy Your Model

This unit advances students into the evaluation and deployment phases of the machine learning lifecycle.

Students learn to assess a model's ability to generalize to new data through various out-of-sample validation

techniques. Students explore different methods to enhance model performance, including the optimization

of hyperparameters and feature selection, using their own implementations along with the built-in capabilities

of scikit-learn.

Students then transition to the deployment of machine learning models and how to make models operational

and available to stakeholders to address real business challenges. This includes discussions on deployment

strategies and the ongoing monitoring of models once they are in production. This unit is designed to provide

students with a comprehensive understanding of how to not only develop but also effectively deploy and

manage machine learning models.

Unit 5 Objectives:

1. Understand the importance of model selection in machine learning

2. Choose model evaluation metrics that are appropriate for the application

3. Choose appropriate model candidates and hyperparameters for testing

4. Set up training/validation/test splits for model selection

5. Apply feature selection techniques to get a better-performing model

6. Explore how to deploy, host, and monitor your model



Unit 6: Special Topics: Ensemble Methods and Unsupervised Learning

This unit covers advanced machine learning techniques: ensemble modeling and unsupervised clustering.

Students will first explore three key ensemble techniques—stacking, bagging, and boosting—examining

algorithms like random forests and gradient boosted decision trees, with opportunities to build and optimize

these models.

The unit then transitions to unsupervised learning, focusing on clustering as the most prevalent technique.

Students will practice implementing clustering using Sci-kit learn, broadening their understanding of machine

learning's applications.

Unit 6 Objectives:

1. Explore the bias-variance tradeoff

2. Improve model performance with ensemble methods

3. Understand the mechanics of three ensemble methods: stacking, random forests and gradient

boosted decision trees

4. Explore unsupervised learning

5. Implement unsupervised clustering



Unit 7: Special Topics: Deep Learning and Computer Vision

This unit equips students with the essential skills and theoretical knowledge needed to apply neural networks

and deep learning techniques effectively. Students begin this unit by exploring the foundational principles of

neural networks. Students are then introduced to the architecture of traditional feedforward neural networks

and are guided through the process of constructing and training these networks using Keras.

The instruction will then extend to deep learning. Students explore applications of deep learning in fields

such as natural language processing and computer vision, highlighting the effectiveness of different neural

network architectures like convolutional neural networks in these areas. The unit’s exercises allow students to

implement a convolutional neural network for image classification tasks.

Unit 7 Objectives:

1. Explore the design of a neural network using the basic components of network architecture

2. See how a neural network is trained and optimized

3. Use Keras to implement a neural network to make predictions

4. Explore the field of Computer Vision

5. Implement a neural network for image classification



Unit 8: Special Topics: Natural Language Processing

In this unit, students explore Natural Language Processing (NLP). The unit covers the NLP pipeline, beginning

with preprocessing and transforming text into numerical features. Students explore various techniques

integral to this initial step and the subsequent stages of the NLP pipeline for effective text data processing.

Students then dive into the utilization of neural network models. They will gain hands-on experience by

implementing a neural network tailored to address an NLP challenge, enhancing their practical skills in

applying complex machine learning models.

This unit’s lab emphasizes proper problem formulation and bias mitigation. They will select from three

datasets previously introduced in the course, preparing and employing these datasets to articulate and solve

a specific machine learning problem. Each student will draft a comprehensive project plan that outlines their

approach to model fitting, evaluation, and iterative improvement, consolidating their learning and preparing

them to tackle real-world NLP problems.

Unit 8 Objectives:

1. Explore the NLP pipeline

2. Use various NLP preprocessing techniques to convert text to data suitable for machine learning

3. Understand how vectorizers are used to convert text into numerical features and how word

4. Explore how word embeddings are used to convert text into numerical features without losing the

underlying semantic meaning

5. Discover how deep neural networks are used in the NLP field

6. Implement a feedforward neural network for sentiment analysis

7. Create and implement a project plan to solve a machine learning problem