We’ll teach you the most in-demand machine learning models and algorithms you’ll need to know to succeed as an MLE. For each model, you will learn how it works conceptually first, then the applied mathematics necessary to implement it, and finally you will get experience training and testing the models. We’ll walk you through the best practices for predictive optimization, like hyperparameter tuning, and how to evaluate your performance. You’ll learn how to pick the right model for the challenge you are facing, and critically, how to implement and deploy these models at scale.

1. Algorithms for both supervised and unsupervised learning

2. Gauging model performance using a variety of cross-validation metrics

3. Using AutoML to generate baseline models

4. Model selection and hyperparameter tuning

5. Bias in models and model drift

6. Deep learning techniques like convolutional, and recurrent neural networks, and generative adversarial networks

7. Recommendation systems

8. Tools: Scikit-Learn, Tensorflow, Pandas, AutoML systems, AWS

Throughout this course, you’ll be introduced to a variety of tools and libraries that are used in both data science and machine learning. These include everything from ML libraries to deployment tools. There will also be refreshers on software engineering best practices and foundational math concepts that every ML Engineer should know.

1. Python Data Science Tools include Pandas, Scikit-learn, Keras, TensorFlow, SQL

2. Machine learning engineering tools including TensorFlow, Flask, AWS, Docker, Kubernetes, FastAPI

3. Software engineering tools including continuous integration, version control with Git, logging, testing, and debugging

4. Working With data pipelines

A critical part of every machine learning engineer’s job is collecting, cleaning, processing, and transforming data. Without quality data, you can’t get quality insights. You’ll learn the best practices and tools for working with data at scale and how to transform a messy, sparse dataset into something worthy of modeling.

1. Exploratory data analysis

2. Cleaning and transforming data for ML systems at scale

3. Working with large data sets in SQL

Machine learning at scale and in production is an entirely different beast than training a model in Jupyter notebook. When you’re working at scale, there are a host of problems that can disrupt your model and its performance. We’ll teach you about the best practices for surmounting these challenges, how to write production-level code, as well as ensuring that you are getting quality data fed into your model.

1. Creating reliable and reproducible data pipelines to ensure your model is well fueled

2. Cloud-based services provided by AWS

3. The machine learning life cycle and challenges that can occur when integrating your model into an application

4. REST APIs, serverless computing, microservices, containerization

Deep learning is a subset of machine learning that uses neural networks with multiple layers to learn and extract complex patterns and representations from data. This advanced machine-learning technique powers many of today’s most cutting edge applications, including generating photorealistic faces of people who have never lived, machine translation, self-driving cars, speech recognition, and more. Deep learning models become more accurate when they are fed more data, so they are excellent for many business problems.

1. Overview of neural networks, backpropagation, and foundational optimization techniques like gradient descent

2. Neural network architectures

3. Transfer learning

4. Training neural networks using Keras and TensorFlow

5. Computer vision including convolutional neural networks, image segmentation, object detection, and generative adversarial networks

6. Natural language processing including large language models, sentiment analysis, and named entity recognition

Ethics and bias in machine learning refer to the principles, guidelines, and considerations surrounding the responsible and fair use of machine learning algorithms and models, ensuring that their deployment and outcomes uphold human values, avoid bias and discrimination, protect privacy, and prioritize transparency and accountability.

1. Algorithmic bias and fairness

2. Privacy concerns in ML

3. Model transparency and interpretability

4. Ethical considerations in ML research and deployment

5. Best practices for responsible AI development and deployment