hi i'm dan and i'm a curriculum writer at code.org i'm here to talk a little bit more about ai bot and how it can help us make decisions with data with a new tool called ai lab most of the time we see examples of artificial intelligence and machine learning at the end of the process after it's already able to make decisions in the last few lessons we've learned more about the start of this process we saw how you can take data investigate it to find patterns and then use those patterns to make decisions we've also learned some of the terms that go along with this process the features are the inputs that we base our decisions on the model is what we create to make decisions and the label is the output from our decisions for small amounts of data human beings are capable of finding these patterns and creating these models just like a machine would but with large amounts of data and with even more features it becomes harder and harder for humans to analyze all the data so we notice patterns sometimes the data can be overwhelming luckily we have machines like ai bot to help us out ai bot can analyze large amounts of data and notice patterns that sometimes humans don't even see using some of the techniques we've seen so far and with a lot more math ai bot can create models and make decisions no matter how much data it looks at this is where machine learning is helpful in solving problems it can analyze large amounts of data and find trends faster than humans can but humans still have an important role to play we help decide what problems ai bot should solve in the first place what data it should use and what features ai bot should pay attention to the predictions that ai bot makes correctly and even those that it makes incorrectly are directly related to our decisions about the data and features ai bot uses and even after ai bot makes its predictions we decide how we want to act on that information as we start looking at more complex problems with large amounts of data we'll need ai bots help to help us work with aibot we'll be using a tool called ai lab ai lab lets you prepare data train a machine learning model and evaluate how that model is doing let's take a look this is ai lab the center of our screen is filled with our data set this is what we'll use to train our model and make a prediction in this example i've loaded our shoe survey data from before where we tried to predict what shoes someone should wear based on several survey questions just above our data set is a sentence starter this sentence represents the model that we're preparing for ai bot the first part's been filled out for us we want to predict what shoe someone should wear based on and then the last part is blank this is because we need to fill this in by selecting different columns we want to use as features in our model when i click on a column the right panel loads with more information about this column i can see a longer description about what this column represents more than just the one or two word heading at the top i can also see a crosstab table this table is helpful for determining whether or not this column will be a useful feature in our model for example i can see that there's a hot spot in my table based on how people want to spend their time folks who want to spend their time in bed are really likely to select sandals as their preferred shoe exploring hotspots in these tables can help determine whether or not this is a useful feature in my model i can also see a bar chart which tells me how often certain values appear in my data even though there's a high relationship between bed and sandals i can see only 18 people said they were going to spend their day in bed if i decide that i want to use this column as one of my features i can press the add feature button i can choose as many columns as i'd like to use as features in my model every time i add a feature my sentence updates to show me what my model will be trained on if i change my mind and want to remove a feature i can press the little x to have it removed from my model once you've explored your columns and selected your features you can press the train button to see how well ai bot can create a model based on these features