Sample Data Science Course Curriculum

gistfile1.md

Sample Data Science Course Curriculum

Unit 1: The Basics

Introduction to Data Exploration

• Describe the data mining workflow and the key traits of a successful data scientist.
• Extract, format, and preprocess data using UNIX command-line tools.
• Explore & visualize data.

Introduction to Machine Learning

• Explain the concepts and applications of supervised & unsupervised learning techniques.
• Describe categorical and continuous feature spaces, including examples and techniques for each.
• Discuss the purpose of machine learning and the interpretation of predictive modeling results.

Unit 2: Fundamental Modeling Techniques

K-Nearest Neighbors Classification

• Describe the setting and goal of a classification task.
• Minimize prediction error using training & test sets, optimize predictive performance using cross-validation.
• Understand the kNN classification algorithm, its intuition and implementation.
• Implement the "hello world" of machine learning (kNN classification of iris dataset).

Naive Bayes Classification

• Outline the basic principles of probability, including conditional probability and Bayes’ theorem.
• Describe inference in the Bayesian setting, including the prior and posterior distributions and the likelihood function.
• Understand the naive Bayes classifier and its assumptions.
• Implement a spam filter using the naive Bayes technique.

Regression and Regularization

• Explain the concepts of regression models, including their assumptions and applications.
• Discuss the motivation for regularization techniques and their use.
• Implement a regularized fit.

Logistic Regression

• Describe the applications of logistic regression to classification problems and probability estimation.
• Introduce the concepts underlying logistic regression, including its relation to other regression models.
• Predict the probability of a user action on a website using logistic regression.

K-Means Clustering

• Explain the purpose of exploratory data analysis, its applications in continuous and categorical feature spaces, and the interpretation and use of clustering results.
• Discuss the importance of the distance function in cluster formation, as well as the importance of scale normalization.
• Implement a k-means clustering algorithm.

Unit 3: Furthering Modeling Techniques

Ensemble Techniques

• Describe general ensemble techniques such as bagging and boosting.
• Build an enhanced classification algorithm using AdaBoost.

Decision Trees and Random Forests

• Describe the use and construction of decision trees for classification tasks.
• Create a random forest model for ensemble classification.

Dimensionality Reduction

• Explain the practical and conceptual difficulties in working with very high-dimensional data.
• Understand the application and use of dimensionality reduction techniques.
• Draw inferences from high-dimensional datasets using principal components analysis.

Recommendation Systems

• Explain the use of recommendation systems, and discuss several familiar examples.
• Understand the underlying concepts, including collaborative & content-based filtering.
• Implement a recommendation system.

Unit 4: other tools

Database Technologies

• Introduce concepts and use of relational databases, alternative database technologies such as NoSQL, and popular examples of each.

Network Analysis

• Describe the use of graphs and graph theory to analyze problems in network analysis.
• Explore network visualization.

Map-Reduce

• Describe the concepts of parallel computing and applications to problems in big data.
• Introduce the map-reduce framework
• Implement and explore examples of map-reduce tasks.

Final Project Working Session

• Where To Go Next
• Review of concepts and examples from preceding weeks.
• Discussion of resources & tools for further study.

Final Project Presentations