Importing Data, Accessing, & Creating a New Experiment | Beginning Azure ML | Part 1

Key Takeaways

Hello internet. Welcome back to our learning Azure ML series. Today we will cover accessing Ashure ML porting importing data into Ashure ML and creating an experiment. If you already know how to do those things, just go ahead and skip to the next video in this series. So how do you access Microsoft Ashure ML? Well, Ashure ML is a web application within Microsoft Ashure itself. So I have Microsoft Azure open and you just click on and access your portal into Microsoft Azure and it should be one of the Azure services that pops up on the left side right so it's actually represented as a beaker that's kind of cute so you click on the machine learning and then access your account and then sign into the ML studio so now we're actually in Microsoft Azure machine learning itself and this is where all of our experiments are kept and as you can see we already have a whole bunch of experiments running. All right, now that we have access ashure ML, let's import some data into Ashure ML. So before we can import data, we have to get data from an external source. And today we'll be using the Titanic data set. It's a very well-known introductory data set. And I'm going to get it from the Keo website cuz Keo is actually hosting a data mining competition right now on the Titanic data set. You can enter it if you want. And we're going to actually download the Titanic training data set uh CSV file. You won't need a kegle account though. But if you don't want to do that, you can just Google Titanic CSV and a whole bunch of hits should come up. All right. So once you have that file downloaded, let's just take a peek inside and see what's in there. So all right, good. So this is a CSV file. It's a flat file meant for transferring a lot of data via text. So now that we actually have data to import into Ashure ML, let's do that. So on the bottom left hand corner of Ashure ML, there is a new button. You want to click on that button and go to data set and go to local file. And then this is where you get to choose your file. Right? So let's import the Titanic CSV. And because I've already imported the Titanic CSV earlier, it asks me if I want to replace the new data set. So this is really handy, right? If you decide that you want to add more data in the future, like if you have an ongoing file that you want to import, and these are all the file types that are supported by Azure ML. Now, also please note that if you're trying to import a spreadsheet, you can just simply import it as a CSV file. That'll be fine. It will read the XLS just fine. And then you want to hit this check box when you're done. All right. So, we have our data and it's somewhere in the Azure machine learning studio. So, let's click on the new button again. Let's create a new experiment. And that'll take us to our workspace where we can actually start building our experiments. Now, we're actually in the machine learning studio. Let's name our experiments, right? So we can have multiple experiments, huge amounts of experiments actually. And let's call this Titanic model. All right. So you'll notice in the background there's a template of how things should look, right? So you have your data, it goes in, it's going into data transformation. Then you have experiment one, experiment two, scoring, validation, etc. So let's go look for our Titanic data that we've imported. So look at this here. So this is all the data that we have access to, right? So there's a whole bunch of uh sample data that comes that sh is shipped with uh uh Azure machine learning and for example you can just double click uh to bring the Titanic data center or you can drag it in right and it will remind you kind of like uh vis. Well that concludes our video. Join us next time when we'll show you how to import non-static data files and data from various external sources. If you like what you just saw, subscribe to our channel or leave us a comment. Let us know if there's a topic you want us to cover and be sure to check us out at data science dojo.com.