YSB 579E - Special Top.in Earth Sys.Sci (Applied Machine Learning in Atmospheric Sciences with Python).

Dersin Amaçları

? Understanding the principal conceptions of machine learning and its algorithms.
? Prepare students for python programming and data analysis, model building, and evaluation in the context of atmospheric datasets.
? Explore various machine learning algorithms and their applications in tasks such as weather forecasting, climate modeling, and environmental monitoring.
? Prepare students for further research or professional work by introducing them to current trends and challenges in the field of atmospheric sciences and machine learning.

Dersin Tanımı

Course Duration: 16 weeks (3 hours for each week, in total 48 hours)
Course Outline
Week 1
? The aims of the course.
? Overview of studies exploring the application of machine learning in atmospheric sciences.

Week 2
1. Python for Data Science (3 hours)
? Introduction to Python libraries: NumPy, pandas, Matplotlib, Seaborn
? Basic data manipulation and visualization (Importing, Summarizing and Visualizing Data)
? Exercises
Week 3
2. Introduction to Machine Learning
? 2.1 Statistical Learning (3 hours)
? Supervised and Unsupervised Learning, Reinforcement
? Training and Test Loss, Validation, Tradeoffs in Statistical Learning, Estimating Risk.
? Exercises
Week 4
? 2.2 Statistical Learning (3 hours)
? Modeling Data
? Multivariate Normal Models
? Normal Linear Models
? Bayesian Learning
Week 5
3. Supervised Learning
? 3.1 Regression Models (3 hours)
? Linear regression, polynomial regression
? Evaluation metrics (MAE, MSE, RMSE)
? Exercises
Week 6
? 3.2 Classification Models (3 hours)
? Logistic regression, decision trees, random forests
? Evaluation metrics (accuracy, precision, recall, F1-score)
Week 7
? 3.3 Model Selection and Tuning (3 hours)
? Hyperparameter tuning (Grid Search, Random Search)
? Cross-validation techniques
? Model comparison and selection
Week 8
Assignments
Week 9
4. Unsupervised Learning
? 4.1 Clustering Algorithms (3 hours)
? Empirical Distribution and Density Estimation
? Clustering via Mixture Models
? Clustering via Vector Quantization
? Exercises
Week 10
? 4.2 Clustering Algorithms (3 hours)
? Hierarchical Clustering
? Principal Component Analysis (PCA)
Week 11
6. Decision Trees and Ensemble Methods
? Top-Down Construction of Decision Trees
? Controlling the Tree Shape
? Bootstrap Aggregation
? Random Forests
? Boosting
? Exercising
Week 12
Assignments
Week 13
Paper – Muhammed Denizoglu and Ismail
7. Machine Learning Workflow (3 hours)
? Data collection and preprocessing
? Model selection and evaluation
? Model deployment and monitoring
? Exercising
Week 14
9. Advanced Topics
? 9.1 Advanced Topics (Deep Learning in Climate and Marine Sciences (3 hours)
Week 15
Presenting the projects
Week 16
Presenting the projects