This is a github repro for the Udacity Nanodegree Programms I haven taken in 2021. It includes all the graded projects. In case you have a question feel free to contact me.
Udacity has a lot of more interessting courses and the options are endless. My goal was to get fast into Data Science and Machine Learning while also understanding the theory behind it. I therefor tried to follow "The Data Science Hierachy of Needs" from bottom to top and selected courses related to the topics. The cumulative effort for the courses is 26 months (based on 10hrs/week).
Note: Udacity reviewers give very useful feedback even after successfully passing a project. I am still implementing most of it.
Recommended time: 5 months, Level: Intermediate, Perequisite: Intermediate Python & SQL
Data Engineering is the foundation for the new world of Big Data. It teaches how to build production-ready data infrastructure, an essential skill for advancing a data career.
- Data Modeling with Postgres
- Data Modeling with Cassandra
- Data Warehouse
- Data Lake
- Data Pipelines with Airflow
Recommended time: 4 months, Level: Intermediate, Perequisite: Python & SQL
Use Python, SQL, and statistics to uncover insights, communicate critical findings, and create data-driven solutions
At the end of the course Udacity recommends to take the Data Scientist Nanodegree Programm.
- Explore Weather Trends
- Investigate a Dataset
- Analyze A/B Test Results (Practical Statistics)
- Wrangle and Analyze Data (Data Wrangling)
- Communicate Data Findings (Data Visualization)
Recommended time: 4 months, Level: Advanced, Perequisite: Python, SQL & Statistics
The Data Scientist Nanodegree program builds on the foundation of the Data Analyst Nanodegree to teach the following skills that companies look for in Data Scientist job candidates:
- Knowledge of how to build supervised machine learning models and making data-driven solutions
- Strong understanding of neural networks, deep learning, and PyTorch
- Ability to build unsupervised machine learning models
- Software engineering and data engineering skills
- Knowledge of how to design experiments and analyze A/B test results.
- Data Science Blog Post
- Disaster Response Pipeline (Data Engineering)
- Recommendations with IBM (Experimental Design & Recommendations)
- Capstone: Dog Project
Recommended time: 3 months, Level: Advanced, Perequisite: Algebra, Calculus, Statistics, & Python
The course is based around the book by Peter Norvig who also gives a lot of the lectures within the programm:
- English: Artificial Intelligence: A Modern Approach
- German: Künstliche Intelligenz: Ein moderner Ansatz
At the end of the course Udacity recommends to take the Deep Learning, Deep Reinforcement Learning and Machine Learning Engineer Nanodegree Programm.
- Build a Sudoku Solver
- Build a Forward-Planning Agent (Automated Planning)
- Build an Adversarial Game Playing Agent (Adversarial Search)
- Part of Speech Tagging (Probabilistic Models)
Recommended time: 4 months, Level: Intermediate, Perequisite: Basic Python
The course recommends the book by Andrew Trask who also gives a lecture on Sentiment Analysis:
- English: Grokking Deep Learning
- German: Neuronale Netze und Deep Learning kapieren: Der einfache Praxiseinstieg mit Beispielen in Python
The Deep Learning Nanodegree program introduces key techniques powering many of the most exciting recent advances in AI. You'll learn about various network architectures and applications, with an emphasis on real-world applications.
- Predicting Bike-Sharing Patterns (Neural Networks)
- Landmark Classification & Tagging for Social Media (Convolutional Neural Networks)
- Generate TV Scripts (Recurrent Neural Networks)
- Generate Faces (Generative Adversarial Networks)
- Deploying a Sentiment Analysis Model (Deploying a Model)
Recommended time: 3 months, Level: Advanced, Perequisite: Experience with Python, Probability, Machine Learning, & Deep Learning.
The course recommends the book by Miguel Morales who also gives a lecture on Actor-Critic Methods:
- English: Grokking Deep Reinforcement Learning
The Deep Reinforcement Learning Nanodegree program teaches how to train AI agents to perform a variety of complex tasks.
- Navigation (Value-Based Methods)
- Continuous Control (Policy-Based Methods)
- Collaboration and Competition (Multi-Agent Reinforcement Learning)
Recommended time: 3 months, Level: Intermediate, Perequisite: Intermediate Python & Machine Learning Algorithms
Learn advanced machine learning techniques and algorithms -- including how to package and deploy your models to a production environment.
- Deploy a Sentiment Analysis Model (Machine Learning in Production)
- Plagiarism Detector (Machine Learning, Case Studies)
- Capstone: Dog Project
At the end of the course Udacity recommends:
- Write a blog post about the course experience or project.
- Apply to at least two jobs with "machine learning" in the description.