Course Agenda
Agenda
Course Agenda
Course led by Nima Safaian, head of trading analytics, Cenovus Energy and Jonathan Regenstein, director of financial services, RStudio, Inc.
08:30 – 09:00
Registration and refreshments
08:30 - 09:00
09:00 – 10:30
Data import / wrangling: working with APIs and data providers, structuring and cleaning data
09:00 - 10:30
- Intro with readr, readxl for csv and excel
- API demo: interface to EIA data using rJSON
- Dplyr for wrangling data
- Lubridate for working with dates
- Interface to SQL and NoSQL databases
10:30 – 10:45
Break
10:30 - 10:45
10:45 – 12:00
Data modelling and visualisation strategies: building and testing algorithms
10:45 - 12:00
- Mechanics of modelling: built in and custom
- Rolling models
- Multiple models: fitting and managing
- Coding algorithmic logic with dplyr
- Ggplot2 and the art of exploratory data visualisation
- Interactive JavaScript visualisations
12:00 – 13:00
Lunch
12:00 - 13:00
13:00 – 14:30
Real world applications
13:00 - 14:30
- Time series analysis – visualising pricing relationship
- Interactive maps using leaflet case study
- Clustering analysis and heatmaps for risk reporting
14:30 – 14:45
Break
10:30 - 10:45
14:45 – 16:15
Shiny and RMarkdown
14:45 - 16:15
- Intro to Shiny: hello world example (app.r and flexdashboard)
- A more complex applications
- Htmlwidgets
- Highcharter
- Reporting with RMarkdown: html and PDF
08:30 – 09:00
Refreshments
08:30 - 09:00
09:00 – 10:30
Forecasting: forecast package, anamolize, prophet, sweep
09:00 - 10:30
- Modelling time series with forecast and anamolize
- Forecasting with the forecast package
- Forecasting with the prophet package
- Tidy and visualise forecasts with sweep
10:30 – 10:45
Break
10:30 - 10:45
10:45 – 12:00
Introduction to machine learning
10:45 - 12:00
- The R landscape for ML
- Model selection
- Cross validation and resampling with rsample
- PCA analysis of commodity prices
- ML time series forecasting: bagging, boosting
12:00 – 13:00
Lunch
12:00 - 13:00
13:00 – 14:30
Machine learning in real world
13:00 - 14:30
- Building an automated insight generator
- Price prediction using logistics regression
- Sentiment analysis using natural language processing and classification methods
14:30 – 14:45
Break
10:30 - 10:45
14:45 – 16:15
Deep learning and high level intro to TensorFlow
14:45 - 16:15
- What is deep learning?
- What is tensorflow?
- Building models using greta package
- A time series forecasting example
16:15 – 23:59
End of course
16:15 - 23:59