hi Matt. I understand that, the more complex the model is, the more accuracy it has. However, in terms of explaining it to an executive level, I see a challenge. Are you including methods to explain your models in this course?
Hi James, Normally we explain what factors contribute to a model. This is called model explainability (interpretability), and we teach this in DS4B 201-R, which is data science for business. We use a tool called LIME for explaining why employees are likely / not likely to churn. Here's a link to the course (which is part of our 5-Course R-Track): university.business-science.io/p/hr201-using-machine-learning-h2o-lime-to-predict-employee-turnover/
@@BusinessScience I think python will really help compete with market leaders like prophet and greykite. IMHO i think your stuff has a lot more to offer than prophet etc, so will be eagerly waiting for the python one.
@@imranh2774 Modeltime integrates all of them, so you get everything in 1. Our philosophy is experiment... Don't go in with any assumptions, let the best algorithms win out. This approach works very well. Will be a year or so for Python. R is available now though.
![Forecasting Multiple Time Series with Modeltime | Bonus Auto-Forecast Shiny App [Lab 46]](/img/n.gif)
Very Nice and Informative... Gr8 work Matt and David .. Kudos to u guys :)
You got it, Vishal!
Very nice course
It sure is!
Hmm - over my head. Looks like a long infomertial. I had hope to see some examples of the use of timetk.
hi Matt. I understand that, the more complex the model is, the more accuracy it has. However, in terms of explaining it to an executive level, I see a challenge. Are you including methods to explain your models in this course?
Hi James, Normally we explain what factors contribute to a model. This is called model explainability (interpretability), and we teach this in DS4B 201-R, which is data science for business. We use a tool called LIME for explaining why employees are likely / not likely to churn. Here's a link to the course (which is part of our 5-Course R-Track): university.business-science.io/p/hr201-using-machine-learning-h2o-lime-to-predict-employee-turnover/
@@BusinessScience Awesome! Thanks Matt.
@@jamespaz4333 You got it!
Hi. Do you have a contact email so I can ask you some questions about the course?
Absolutely. You can reach us through our contact form. www.business-science.io/contact.html
Is python version of the modeltime available ?
Not yet... Took over a year to develop in R. Will take some time for Python.
@@BusinessScience I think python will really help compete with market leaders like prophet and greykite. IMHO i think your stuff has a lot more to offer than prophet etc, so will be eagerly waiting for the python one.
@@imranh2774 Modeltime integrates all of them, so you get everything in 1. Our philosophy is experiment... Don't go in with any assumptions, let the best algorithms win out. This approach works very well. Will be a year or so for Python. R is available now though.
Will it be feasible for intermittent demand?
Yes absolutely! Intermittent is no problem. You just need to learn how.