#Management and Analysis of Physics Datasets @UniPD Role: Teaching Assistant Lectures for laboratorory of Management and Analysis of Physics Datasets course of Msc in Physic of Data @UniPD 2018-2019

Lecture 1

• Slides on Dask and and Clsuter / HPC
• Principles of distribute calculos with Dask
• Principle of parallel calculos with Dask
• Homework 1
• Example of distributed Cross validation with Keras
• Solutions

Lecture 2

• Slides on Dask and and Clsuter / HPC
• Distribute calculos with Dask
• Parallel calculos with Dask
• Example of distributed Cross validation with Keras
• Solutions
• Example of distributed ML

Lecture 3

• Distributed dataset
• Distributed remote dataset
• CSV vs H5 dataset
• Dask Dataframe
• Concepts of map reduces
• Homework 2

Lecture 4 / 5

• Eample of Simple map reduce
• Distributed Machine learning
• Incremental learning
• Parallel Post fit
• Distributed Tesnforflow and Keras over CPU/GPU cluster
• Solutions
• Ditributed regression