Udemy Data Science in Layman's Terms: Time Series Analysis

Modeling Time Series Data

What you'll learn​

• Time series forecasting with modern nonlinear models, neural networks, and AI
• Time series classification, with a project on predicting heart attackes from ECG data
• Time series segmentation, with a project categorizing distinct periods of football QB performance
• Signal processing, with a project detecting gravitational waves hidden amongst noise
• Anomaly detection, with a project detecting faulty inverters at solar power plants
• Geospatial-temporal analysis, with a project creating a dashboard to analyze crime in San Francisco
• How to build a dashboard with Dash and Plotly
• How to deploy machine learning as a service (MLaaS), using an API
• How to generate music with AI
• How to build & utilize custom neural networks for time series, including LSTMs and Transformers

Requirements​

• Basic knowledge of math and statistics
• Python (only required for the projects)
• Knowledge of machine learning and neural networks would be helpful

Description​

This course explores a specific domain of data science: time series analysis. The lectures explain topics in time series from a high level perspective, so that you can get a logical understanding of the concepts without getting intimidated by the math or programming. Whether you are new to time series or an experienced data scientist, this course covers every aspect of time series. Topics in time series analysis include:

• Forecasting - Predicting the future
• Classification - Categorize a series
• Segmentation - Breaking a series into periods of distinct characteristics
• Anomaly Detection - Identifying unexpected observations
• Signal Processing - Extracting signal from noise
• Geospatial-Temporal Analysis - Analyzing time series with a location component

The later half of the course entails several projects for you to get your hands dirty with time series analysis in Python. You will learn about modern time series forecasting models and AI, how to build them, and implement them to do extraordinary things.

• Generate music with AI
• Deploy a model to an API to provide machine learning as a service (MLaaS)
• Build a dashboard with Dash/Plotly
• Build different types of RNNs and Transformers, using TensorFlow, for time series modeling
• Analyze different types of data sources, like CSV, JSON, GeoJSON, HDF5, and MIDI

By the end of this course, you will be able to handle any time series problem. You will be equipped with the knowledge to build powerful forecasting models, and be able to deploy them.

Who this course is for:​

• Data Scientists who want to practice time series problems with Python
• Anyone who want to learn more about time series