Data Science Immersive

Dive into a series of self-paced lessons on the essentials of Python programming and applied math for data science before the course begins.

• Explore fundamental Python programming
concepts, including variables, lists, loops,
dictionaries, and data sets.
• Leverage programming tools like GitHub and
the command line interface to manage data
science projects.
• Practice solving coding challenges similar
to the questions used in task-based data
science interviews.
• Write and run Python functions using
multiple arguments.
• Discover how key math concepts like statistical
significance and probability distribution are
applied throughout data science.

Get acquainted with essential data science tools and techniques, working in a programming environment to gather, organize, and share projects and data with Git and UNIX.

• Demonstrate familiarity with introductory
programming concepts using Python and NumPy
to navigate data sources and collections.
• Utilize UNIX commands to navigate file systems
and modify files.
• Learn to track changes and iterations using Git
version control from your terminal.
• Define and apply descriptive statistical
fundamentals to sample data sets.
• Practice plotting and visualizing data using Python
libraries like Matplotlib and Seaborn.

Project: Apply NumPy and Python programming skills
to answer questions based on a clean data set.

Perform exploratory data analysis. Generate visual and statistical analyses, using Python and its associated libraries and tools to approach problems in fields like finance, marketing, and public policy.

• Design an experimental study with a well-thought-
out problem statement and data framework
• Use Pandas to read, clean, parse, and plot data,
extracting and rearranging data through indexing,
grouping, and JOINing.
• Review statistical testing concepts (p values,
confidence intervals, lambda functions,
correlation/causation) with SciPy and StatsModels.
• Learn to scrape website data using popular
scraping tools.
• Explore bootstrapping, Resampling and building
inferences about your data.

Project: Leverage Pandas to apply advanced NumPy
and Python skills cleaning, analyzing, and testing data
from multiple messy data sets.

Explore effective study design and model evaluation and optimization, implementing linear and logistic regression, and classification models. Collect and connect external data to add nuance to your models using web scraping and APIs.

• Use scikit-learn and StatsModels to run linear
and logistic regression models and learn to
evaluate model fit.
• Begin to look at classification models by
implementing the k-nearest neighbors (kNN)
algorithm.
• Articulate the bias-variance trade-off as you
practice evaluating classical statistical models.
• Use feature selection to deepen your knowledge of
study design and model evaluation.
• Learn to apply optimization and regularization for
fitting and tuning models.
• Dive into the math and theory behind how gradient
descent helps to optimize loss functions for
machine learning models.

Project: Explore, clean, and model data based on
a provided data set, outlining your strategy and
explaining your results.

Build machine learning models. Explore the differences between supervised and unsupervised learning via clustering, natural language processing, and neural networks.

• Define clustering and its advantages
and disadvantages as compared to
classification models.
• Build and evaluate ensemble models using decision
trees, random forests, bagging, and boosting.
• Get acquainted with natural language processing
(NLP) through sentiment analysis of scraped
website data.
• Learn how Naive Bayes can simplify the process of
analyzing data for supervised learning algorithms.
• Explore the history and use of Hadoop, as well as
the advantages and disadvantages of using parallel
or distributed systems to store, access, and analyze
big data.
• Understand how Hive interacts with Hadoop
and discover Spark’s advantages through big data
case studies.
• Analyze and model time series data using the
ARIMA model.

Project: Students will scrape and model their own data
using multiple methods, outlining their approach and
evaluating any risks or limitations.

Dive deeper into recommender systems, neural networks, and computer vision models, implementing what you’ve learned to productize models.

• Compare and contrast different types of neural
networks and demonstrate how they are fit with
back propagation.
• Build and apply basic recommender systems in
order to predict on sample user data.
• Work with career coaches to create and polish
your professional portfolio.
• Practice with data science case studies to prepare
for job interviews.

Project: Choose a data set to explore and model,
providing detailed notebook of your technical
approach and a public presentation on your findings.