Artificial Intelligence Science Lab
The collection, representation and interpretation of data are everyday tasks for students of all ages. The science subjects are a rich source of data, whether students measure plant growth in biology, the time it takes for a chemical reaction to occur in Chemistry, or how far a projectile will fly in physics.
Common to these and many other experiments is that we want students to understand the data based on their interpretation of input-output relations and whether these relations are simply linear or more complex.
Those of us with experience in the matter can often tell by just looking at the charts how these relations are. But learners who take their first steps in data science often struggle. This is particularly true when multiple variables affect the outcome.
Artificial Intelligence is particularly suitable for looking at datasets and learning the relations between input variables and outputs. Because programming can get in the way when dealing with AI libraries, we have made a simple toolkit that students can use without writing code. We call it the AI Science Lab. The code-freeness makes the lab suitable for all students who understand what a table is.
Let's take a look:
Students enter the data from their science experiments in the table on the left-hand side. As they enter it, the data gets plotted in the charts. At the top of the screen is a simple deep-learning artificial neural network. Students scroll through the data with the input sliders. The instructions on the right-hand side menu help students to navigate.
The input table allows students to enter records, rename the columns, de-select them if their experiment has fewer than three input variables, etc. In addition, students can store their data on the server to avoid losing it between lessons.
Training the AI:
When the data has been entered, the real fun begins. By clicking on the Start learning button at the bottom of the screen, the AI starts to learn about the data.
First, it tries to recreate the relations between inputs and output. Red lines on the chart appear to inform the trends that the AI sees in the data.
Is the output growing with the input? Is it shrinking? How much growth is there? The red curves provide important cues concerning the nature of the relationship between input and output.
Using the AI Model
Students use the input sliders to test their model once the training process is complete (or stopped manually). Then, they compare the calculated output from the AI against the results they would expect. The AI Science Lab supports both interpolation and extrapolation to explore data areas between our outside the records students collected previously.
You find the AI Science Lab directly here of by navigating from the My Computer Brain Website to the Artificial Intelligence (Secondary+Tertiary) course. Look out for the yellow robot. Teacher accounts are free and student accounts are very reasonably priced. Check out the getting started blog post.
The AI Science Lab is the last activity in the AI course.
We hope you and your students like the new Lab. Please ping us if we can be of any help !