Building dashboards in Python and Continuous Integration of Machine Learning

Machine Learning Meetup Prague
Machine Learning Meetup Prague
Public group
Location image of event venue


Dashboards in pure Python with a dash of Plotly - Anastázie Sedláková
In data analysis, we often do predictions to make decisions. They say that "A picture is worth a thousand words". But how to make this picture easy digestible to various people in the company (YES, I do not mean matplotlib plots in Jupyter Notebook)?
I will show examples how you can create interactive plots with few lines of code in Python (using Plotly library) and then gather them into the dashboard (again in Python using Dash library). You will see how easy it is to make visualizations on map. If nothing else, you will see plenty of nice colorful images.
Dash is an analogy of R's Shiny package and makes it super easy for data scientists to create web application without knowing HTML, CSS and JavaScript.

Continuous Integration of Machine Learning into Production Processes - Filip Sedlák
In Twisto, Risk analysts, who train neural networks and prepare rules
for analytics and alerting, are a separate team from the application
developers. What needs to happen before a newly researched model gets to
score real-life purchases or before an improvement in behavioral
analytics assesses real customers' behaviour? We prefer to minimize the
time and developers' involvement for such changes.

Language: English (will be specified)

Anastázie (Nasťa) Sedláková
Anastázie (Nasťa) Sedláková is doing data analysis (with big passion for visualization), co-organizing programming courses (mostly Python) and raising tiny zen-master (a.k.a toddler).
Nasťa did her PhD in the field of statistical genetics and worked few years as bioinformatician. She then moved outside of biology and academia into financial world. She is currently working on data analysis projects mainly in Twisto, but also for other companies (e.g Behavio). Her latest work (apart from doing various workarounds when dev doesn't have time) is analyzing marketing campaigns.
Second big passion of Nasťa is teaching adult people to program with strong focus on applying acquired knowledge in their professional field. After teaching few years for various non/profit organizations and Charles University, she is now co-organizing courses with her husband.

Filip Sedlák
Filip Sedlák is a DevOps consultant with experience in Machine Learning.
He likes to work with startups and design data-heavy systems. Filip
studied cheminformatics, worked on chemo- and bio-informatics solutions for pharmaceutical companies. He was a co-founder of NeuronSW, a machine learning startup that predicts failures of expensive machinery by sound. At the moment, Filip works mostly for Twisto, a fin-tech company.
Twitter: @koraalkrabba

- 17:45 - 18:00 - Your arrival
- 18:00 - 18:40 - Dashboards in pure Python with a dash of Plotly - Anastázie Sedláková
- 18:40 - 18:50 - Short break
- 18:50 - 19:30 - Continuous Integration of Machine Learning into Production Processes - Filip Sedlák
- 19:30 - 21:00 - Networking in Bitcoin Coffee