Introduction to Data Science

Objectives

1. Recognize the main concepts and methodologies in Data Science.
2. Know and apply simple supervised and unsupervised methods.
3. Create and use decision trees and artificial neural network models.
4. Use and interpret evaluation measures for supervised and unsupervised models.
5. Acquire skills in the use of Python's scienceKitLearn and Pandas libraries in Data Science applications applied to ocean data.

Program

1. Decision Support Systems (DSS). Big Data and Data Mining Technologies.
2. The CRISP-DM data science methodology, domain knowledge, exploration, and pre-processing of data about the ocean.
3. Unsupervised learning: cluster analysis and dimensionality reduction.
4. Supervised learning. Regression and classification. Naïfe Bayes, Nearest Neighbors, and Linear models.
5. Use and interpret evaluation measures of supervised and unsupervised models.
6. Induction of classification, regression, and model trees: the problem of overfitting.
7. The Linear Perceptron. Multilayer perceptron networks for regression and classification.

Teaching Methodologies

The theoretical-practical classes include expository segments in which the concepts are presented using examples and demonstrations that illustrate the use of the algorithms and concepts. Some dynamic methodologies based on gamification are used.

The more practical parts are filled with the joint resolution of worksheet activities.

Assessment by frequency - frequency (NF) and a group project (PG) and weekly asynchronous activities (AA).

The final grade (CF) is calculated through:

If NF >=7 then CF=0.50*NF + 0.35*PG + 0.15*AA if not CF=NF

Evaluation by exam: project (PG) and written exam (NE).

The final grade (CF) is calculated according to:

If NE>=10 then CF= 0.6*NE + 0.4*PG if not CF=NE

Bibliography

- Aurélien Géron. (2022). Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow (3.ª ed.). O’Reilly Media.
- Jake VanderPlas. (2022). Python Data Science Handbook. O’Reilly Media (2.ª ed.).
- Andriy Burkov (2019) The Hundred-Page Machine Learning Book. Author Edition
- Oliver Theobald (2021) Machine Learning for Absolute Beginners: A Plain English Introduction. Scatterplot Press.