Full Stack Dad
.Net developer working across banks, wealth management, insurance, transport, media, lenders, startups, myself
Currently an Engineering Manager @ Bankwest.
Not a data scientist - just an enthusiast
Machine Learning 101
Machine learning gives computers the ability to learn without being explicitly programmed (Arthur Samuel, 1959).
Shallow vs Deep
Tools of choice for ML
What is ML.Net?
OS machine learning framework for .Net developers
Originally developed by MSFT Research
Used in Windows, Bing, PowerPoint, Excel & more
Current preview release is 0.4 (not officially released yet)
Initial release --> 7th May 2018
Why use ML.Net?
Leverage C#/F# capability
Output is .zip file --> easy deployments
C#/F# samples range from basic to end to end apps
eShopDashboardML is a web app with Product Sales Forecast predictions using
What can it do?
Their experiments show faster& more accurate than xgboost
From Microsoft Distributed Machine Learning Toolkit
Microsoft Cognitive Toolkit [todo]
My bank does not give me an overview of spending categories
My tech itch
Build a model that can predict a category using the transaction description
- Retrieve and clean the data
- Decide what algorithm to use
Create a ‘Learning Pipeline’ to iterate over the model build process
- Evaluating result metrics
- Exploring the output model file
Retrieve the data
Decide what model?
Multiclass classification is the problem of classifying instances into one of three or more classes
|Banking||Entertainment||Food & Drinks|
|Groceries||Health & Beauty||Income|
Loading the data
Transforming the data
Configuring a learning model
Training the model
Using the trained model (e.g. to obtain predictions)
Code time finally...
Stochastic Dual Coordinate Ascent
Fast GBM --> WINNER
Predicting BS using ML.Net
By chris mckelt