To finish off the machine learning topic, I spent some time experimenting with how my neural network handled images containing multiple labels. I tested it with images that contained combinations of the categories we had been working with: trees, windows, and seating. Interestingly, I didn’t notice a pattern in how the network classified these images, and the differences in probability between the different labels weren’t very telling either. This got me thinking about how I could possibly build a network that could recognise multiple labels in an image. If I have the time, I’d like to explore this further. I find these kinds of concepts fascinating, and I’m interested in understanding how neural networks determine and weigh probabilities at a deeper level.

This week we were introduced to the basics of Arduino boards and circuitry. I had barely done circuitry before so I struggled more with building it than writing any of the code. But I still ended up succeeding to make my LED flash on and off.

Using the Arduino boards reminded me of being shown how to use Raspberry Pis when I was in primary school. Although I don’t have much memory of how they actually work, I remember being really excited about it. So, delving into circuitry on this topic has inspired me to find my old Raspberry Pi and start experimenting with it again.

Last week, I had some anxiety about the industry folder and was finding it quite overwhelming, but I switched to researching and looking for real jobs available right now, rather than looking at companies. Doing this made me feel much calmer about the future, as I saw that there are, in fact, many different types of opportunities out there that I can imagine myself doing.