Streamline Your Animal Science Sales Forecast Automation with airSlate SignNow

How to create outlook signature

in this video we're going to look at how to use automl specifically profit from Facebook to do a lot of what we were doing in the previous videos completely automatically let's real quick take a look at the time series demand forecasting data set that we're going to use this data set is hosted on kaggle and I created it it's essentially a simulation that looks at forecasting the demand of six restaurants in a beachfront setting you've got time series data natural language processing data and computer vision data let's have a look at all three and this data set I have a link to it in the description this data set exhibits seasonality and Trend both so that's something that you have to be aware of as you're trying to forecast into the future seasonality is the fact that you're seeing this go up and down based on the month of the year and if you zoom in further you'll see that there's even seasonality by the week Trend refers to the this whole thing is increase gradually over time especially if you look at the Peaks you have a bunch of different items that you are trying to forecast the sales for and you've got historic data as it goes into the future there you see a bit better kind of by week you can see this product was clearly discontinued there uh but you you need to forecast when a product discontinues what what is that going to do to the rest is it going to cause other ones to fill in the gap or will that demand simply go away the files are here the primary one that you're going to deal with is this one called sales train which is your sales over time you can see the dates here the item which are those items that we were just looking at the price that it was sold for and how many items it was sold for that day these are all of the days and the items are unique to each restaurant you don't have multiple restaurants selling the same item they're very similar across some of the restaurants the actual items are here you have some information about them they're in tabular form each item is sold by a particular store or restaurant store ID and the restaurants are here for natural language processing I recommend doing something maybe not with the restaurant names because there's not that many of them but the item names you could certainly use natural language processing to maybe extract some further information there's also computer vision which are these pictures that were taken at the street where the five restaurants are at showing the number of people there so you could use something like yellow or other deep learning packages computer vision packages to count how many people both are on the beach and on the street because those tell you different things I did run a kaggle competition with this data set and some of my students at washu you can see the root mean Square errors that these teams were able to accomplish and some of their code is in the code tab I'll also put a link to the kaggle competition that I ran kaggle Community competition so the code is here on the kaggle website that corresponds to the data set because I've uploaded it to kaggle you can run these right in kaggle and that works really pretty well here I load in the CSV files just like before and what I'm going to do is I'm going to combine all of the items together because profit is set up so that it does not handle multiple items so the 100 items that we have here we would end up creating 100 different profit models to do that but I'm just going to look at it this way uh as all of the items sum together so in aggregate how many items are we selling you could certainly break it apart and get profit trying to predict the individual items separately so to make use of it it's really very easy you set a profit and then you fit the model to the data frame and you're telling it which columns you are particularly interested in so we're we're going to be predicting the the number of items by date and you can then request that it make future predictions to do that you create a date a future data frame so an entire year into the future using their helper function there and then you can display the the predictions it shows you the date that they're going that they're at in the future and then the Y hat that's the prediction and then the lower and upper so kind of giving you a cone of of uncertainty and this is what it looks like to the left these are the three years of data that I gave you as training and you can see it's it's figured out the seasonality uh it's it's maybe figured out some of the trend that these two peaks look pretty pretty similar because as you can see there is a trend in here it's increasing especially if you look at the heights of those Peaks this one looks at the same level so I'm I'm not seeing that it's figured out the trend nearly as well and then you can also look at sort of the cone of uncertainty the further you get into the future the less sure that it that it is and it's detecting it is detecting the weekly seasonality so that's quite cool because it is detecting the Friday and Saturday overall you tend to have more and it's also detecting the yearly seasonality that in the summer we tend to have more sales so this is profit which is getting a lot of traction in the kaggle community and it is meta's attempt at providing automl for time series prediction certainly worth taking a look at thank you for watching this video and if you'd like to see more on the series check out the playlist where I've done a number of presentations and videos on demand forecasting based on a two-day course that I put together earlier this year