How to train ML

8.45 am to 12.30 pm CT,
30 July 2021

We’re Finally Here. Training ML Models.

What We’ve Been Waiting For.

Set Up Entire Modelling Processes To

Maximize Performance And Prevent Overfitting.

Swap Algorithms In And Out To Find

The Best Parameters For Each One.

Pro data scientists  spend quality time in the steps leading up to training ML models.

Exploring the data​, Cleaning the data​, Engineering new features​. They do this because improved data defeats fancier algorithms.

Split datasets

Cross-Validation

Hyperparameters

Fit and tune models

Select winning models

Amazing Partners & Sponsors

Agenda

30 July 2021 (8.45 am to 12.30 pm CT)
08:45 am
Aganda 1

Introduction of Speakers

-by Nitin Naveen, Vice president-Innovation Strategy, AIhubspot

 

09:00 am
Aganda 2

Split Datasets

-by Michael Felicity

 

09:20 am
Aganda 4

What are hyperparameters

by Anna Upshur

 

09:40 am
Aganda 5

Model parameters vs hyperparameters

by Kimchi Son-ming-yin

 

10:00 am
Aganda 6

Cross Validation concepts

by Alex Castillo

 

10:20 am
Aganda 8

Break

 

10:35 am
Aganda 7

Fitting and tuning models

– by Anna Upshur​

 

 

11:15 am
Aganda 6

Selecting winning models

by Joanne Vaniccola

 

11:35 am
Aganda 6

Workshop on split datasets / Cross Validation

 ​

 

 

11:55 am
Aganda 6

Workshop on fit and tune / selecting winning models

 

 

 

12:15 pm
Aganda 7

Q&A