Business Science’s Time Series Course is Incredible
I’m a time series fan. Big fan. My first job out of grad school was for a utility company building econometric time series analysis and forecasting models. Lots of ARIMAs and neural nets. However, that was now over 10 years ago (don’t know how the hell that happened).
This post contains affiliate links that help to offset the cost of running the blog, plus the link gives you a special 15% discount. If you use the link, thank you!
I’m a time series fan. Big fan. My first job out of grad school was for a utility company building econometric time series analysis and forecasting models. Lots of ARIMAs and neural nets. However, that was now over 10 years ago (don’t know how the hell that happened).
In almost every position I've held in data, a question has come up that involved a time series (not a surprise that business cares about what has happened over time). Often, I was the only one who had any knowledge of time series on my team. I'm not sure why it isn't taught as a standard part of most university programs that are training data scientists, but it's just unfortunately not. I believe that understanding time series analysis is currently a great way to differentiate yourself, since many in the field are just not well versed in it.
I wanted to understand what was current in the world of applying time series analysis to business. It had been a real long time since I had given the subject some of the love and attention, and I thought taking this Business Science course would be the perfect way to do that.
My History With Business Science Courses:
I’ve previously written about Business Science’s first course, you can check it out here. I've also taken his first Shiny app course (there’s a more advanced one as well) and went from zero to Shiny app in 2 days using survey data I collected with Kate Strachnyi. It was a real win.
The app is still on my site here, just scroll down. For this little flexdashboard app I went from basically zero Shiny to having something that was useful in 2 days leveraging only the first 25% of the course. The course cannot actually be completed in 2 days. It's also worth noting that the course builds an app with much more functionality than mine. It’s a long course.
Back to the Time Series Review:
It’s broken into three different section:
Things I freakin’ love
The sexy
Everything else
Things I freakin’ love:
You’re learning about packages from the package creator. Who is going to understand a library better than the person who wrote it?. Matt built both modeltime and timetk that are used in this course. I find that super impressive. These packages are also a step up from what was currently out there from a "not needing a million packages to do what I want" perspective.
He uses his own (anonymized) data fromBusiness Science to demonstrate some of the models. I haven’t seen others do this, and I think it’s cool. It’s a real, practical dataset of his Google Analytics and Mailchimp email data with an explanation of the fields. If you don’t have analytics experience in e-commerce and are thinking about taking a role in e-commerce, definitely give some thought to this course.
I love how in-depth he gets with the subject. If you follow all that is covered in the course, you should be able to apply time series to your own data.
The Sexy:
Ok, so I’m sure some are interested in seeing just how “cutting edge” the course gets.
Once you're combining deep learning Gluon models and machine learning models using ensembling methods, you might be the coolest kid at work (but I’m not making any promises). Gluon is a package that was created by Amazon in Python. So you’ll leverage both Python and R for Gluon.
Some of the deep learning algorithms you’ll learn how to leverage are:
DeepAR
DeepVAR
N-Beats
Deep Factor Estimator
Module 18 of the course is where you'll get into deep learning. A couple years ago I might have said "deep learning, bah humbug, requires too much computing power and isn't necessary, simpler is better." As things change and progress (and computers get even more beefy) I'm definitely changing my tune. Especially as an ensemble N-Beats algorithm beat the ES-RNN's score in the M4 competition. M competitions are prestigious forecasting challenges, and they've historically been won by statistical algorithms. (I wouldn't have known this information without this course). The stuff being taught in this course is very current and the sexy new techniques that are winning the big competitions.
Here's a look at the syllabus for preparing the data and learning about the DeepAR model. You're doing log transformations, Fourier Series, and when you get to modeling the course even covers how to handle errors. I just love it. I know I'll be referring back to the course when a time series use case pops up in the future.
The course covers 17 different algorithms. I'm trying to think if I could name 17 algorithms off the top of my head… it’d take me a minute. ARIMA is obviously included, because It’s like the linear regression of time series. You’ll go through ARIMA, TBATS (a fave because you don’t need to worry about stationarity the way you do with ARIMA. I’ve used this one in industry as well).
Along with these other algos:
ARIMA Boost
Prophet Boost
Cubist
KNN
MARS
Seasonal decomposition models
Then you’ve got your ensemble algos being leveraged for time series:
GLMNET
Random Forest
Neural Net
Cubist
SVM
Strap in for 8 solid hours of modeling, hyperparameter tuning, visualizing output, cross-validation and stacking!
Everything else:
Matt (the owner of Business Science) speaks clearly and is easy to understand. Occasionally I'll put him on 1.25x speed.
His courses in general spend a good amount of time setting the stage for the course. Once you start coding, you’ll have a great understanding of where you’re going, goals, and context (and your file management will be top notch), but if you’re itching to put your fingers on the keyboard immediately, you’ll need to calm the ants in your pants. It is a thorough start.
You have to already feel comfy in R AND the tidyverse. Otherwise you’ll need to get up to speed first and Business Science has a group of courses to help you do that. You can see what's included here.
Before we finish off this article, one super unique part of the course I enjoyed was where Matt compared the top 4 time series Kaggle competitions and dissected what went into each of the winning models. I found the whole breakdown fascinating, and thought it added wonderful beginning context for the course.
In the 2014 Walmart Challenge, taking into account the “special event” of a shift in holiday sales was what landed 1st place. So you're actually seeing practical use cases for many of the topics taught in the course and this certainly helps with retention of the material.
Likewise, special events got me good in 2011. I was modeling and forecasting gas and the actual consumption of gas and number of customers was going through the roof! Eventually we realized it was that the price of oil had gotten so high that people were converting to gas, but that one tripped me up for a couple months. Thinking about current events is so important in time series analysis and we'll see it time and again. I've said it before, but Business Science courses are just so practical.
Summary:
If you do take this course, you’ll be prepared to implement time series analysis to time series that you encounter in the real world. I've always found time series analysis useful at different points in my career, even when the job description did not explicitly call for knowledge of time series.
As you saw from the prerequisites, you need to already know R for this course. Luckily, Business Science has created a bundle at a discounted price so that you can both learn R, a whole lot of machine learning, and then dive into time series. Plus you’ll get an additional 15% off the already discounted price with this link. If you're already comfortable in R and you're just looking to take the time series course, you can get 15% off of the single course here.
Edit: People have asked for a coupon to buy all 5 courses at once. That's something I'm able to do! Learn R, machine learning, beginner and advanced Shiny app development and time series here.
Effective Data Science Presentations
If you're new to the field of Data Science, I wanted to offer some tips on how to transition from presentations you gave in academia to creating effective presentations for industry.Unfortunately, if your background is of the math, stats, or computer science variety, no one probably prepared you for creating an awesome data science presentations in industry. And the truth is, it takes practice. In academia, we share tables of t-stats and p-values and talk heavily about mathematical formulas. That is basically the opposite of what you'd want to do when presenting to a non-technical audience.If your audience is full of a bunch of STEM PhD's then have at it, but in many instances we need to adjust the way we think about presenting our technical material.I could go on and on forever about this topic, but here we'll cover:
Talking about model output without talking about the model
Painting the picture using actual customers or inputs
Putting in the Time to Tell the Story
Talking about model output without talking about the modelCertain models really lend themselves well to this. Logistic regression, decision trees, they're just screaming to be brought to life.You don't want to be copy/pasting model output into your data science presentations. You also don't want to be formatting the output into a nice table and pasting it into your presentation. You want to tell the story and log odds certainly are not going to tell the story for your stakeholders.A good first step for a logistic regression model would just be to exponentiate the log odds so that you're at least dealing in terms of odds. Since this output is multiplicative, you can say:"For each unit increase of [variable] we expect to see a lift of x% on average with everything else held constant."So instead of talking about technical aspects of the model, we're just talking about how the different drivers effect the output.
We could, however, take this one step further.
Using Actual Customers to Paint the Picture: I love using real-life use cases to demonstrate how the model is working. Above we see something similar to what I presented when talking about my seasonality model. Of course I changed his name for this post, but in the presentation I would talk about this person's business, why it's seasonal, show the obvious seasonal pattern, and let them know that the model classified this person as seasonal. I'm not talking about fourier transforms, I'm describing how real people are being categorized and how we might want to think about marketing to them. Digging in deep like this also helps me to better understand the big picture of what is going on. We all know that when we dig deeper we see some crazy behavioral patterns.Pulling specific customers/use cases works for other types of models as well. You built a retention model? Choose a couple people with a high probability of churning, and a couple with a low probability of churning and talk about those people."Mary here has been a customer for a long time, but she has been less engaged recently and hasn't done x, y, or z (model drivers), so the probability of her cancelling her subscription is high, even though customers with longer tenure are usually less likely to leave.
Putting in the Time to Tell the Story: As stated before, it takes some extra work to put these things together. Another great example is in cluster analysis. You could create a slide for each attribute, but then people would need to comb through multiple slides to figure out WHO cluster 1 really is vs. cluster 2, etc. You want to aggregate all of this information for your consumer. And I'm not above coming up with cheesy names for my segments, it just comes with the territory :).It's worth noting here that if I didn't aggregate all this information by cluster, I also wouldn't be able to speak at a high level about who was actually getting into these different clusters. That would be a large miss on my behalf, because at the end of the day, your stakeholders want to understand the big picture of these clusters.Every analysis I present I spend time thinking about what the appropriate flow should be for the story the data can tell.
I might need additional information like market penetration by geography, (or anything, the possibilities are endless). The number of small businesses by geography may not have been something I had in my model, but with a little google search I can find it. Put in the little extra work to do the calculation for market penetration, and then create a map and use this information to further support my story. Or maybe I learn that market penetration doesn't support my story and I need to do more analysis to get to the real heart of what is going on. We're detectives. And we're not just dealing with the data that is actually in the model. We're trying to explore anything that might give interesting insight and help to tell the story. Also, if you're doing the extra work and find your story is invalidated, you just saved yourself some heartache. It's way worse when you present first, and then later realize your conclusions were off. womp womp.
Closing comments: Before you start building a model, you were making sure that the output would be actionable, right? At the end of your presentation you certainly want to speak to next steps on how your model can be used and add value whether that's coming up with ideas on how you can communicate with customers in a new way that you think they'll respond to, reduce retention, increase acquisition, etc. But spell it out. Spend the time to come up with specific examples of how someone could use this output.I'd also like to mention that learning best practices for creating great visualizations will help you immensely.
There are two articles by Kate Strachnyi that cover pieces of this topic. You can find those articles here and here. If you create a slide and have trouble finding what the "so what?" is of the slide, it probably belongs in the appendix. When you're creating the first couple decks of your career it might crush you to not include a slide that you spent a lot of time on, but if it doesn't add something interesting, unfortunately that slide belongs in the appendix.I hope you found at least one tip in this article that you'll be able to apply to your next data science presentation. If I can help just one person create a kick-ass presentation, it'll be worth it.