Week 21

Lecture

This week, we had a guest lecturer who works as an AI consultant and runs her own business. This was a good introduction to industry AI. Dr Bear shared her experience of applying machine learning methods to solve both research and business problems, giving us real-world insight into how these technologies are used beyond the classroom. Her passion for data began during her transition between education and work, and she spoke about how that interest evolved into a career in AI and consultancy.

One of the key takeaways from her talk was the importance of abstraction when tackling data-related challenges. Being able to step back and see the bigger picture can often lead to more effective solutions. She also encouraged us to remain open to opportunities and not to hesitate to ask for what we want, whether that’s mentorship, project involvement, or career advice.

She also spoke about a current project she’s working on, which involves studying radicalisation in online conversations. Her team is using machine learning to explore questions about how radical ideas spread online and how certain groups can be better protected. The data they’ve gathered is helping to inform strategies for addressing these complex issues, showing how AI can be used for meaningful, socially impactful work. Her talk was inspiring and provided a powerful example of how technical skills can be applied to real-world problems with lasting significance.

Workshop

This week marked an important step forward in my understanding of machine learning, as we were officially in traduced to the task for the Machine Learning Project. We began by outlining what the project would achieve and identifying the key components needed to bring it to life. With that foundation in place, we started learning the basics of machine learning, including how data is used to train models and make predictions. As part of this learning process, I coded my very first machine learning script. It was a valuable introduction to the logic and structure behind ML programming, and it helped me understand how models can begin to recognize patterns from data. In the introduction we also looked at weights and how they can be calculated and added into equations to predict.  

In addition to starting the project, we had our third “Need to Know” session of the year. We then had more workshops on Thursday that built on the work we began earlier in the week, particularly focusing on how to adapt and customize the ML script using different datasets. I experimented with replacing the original data with new, tailored inputs to see how the model’s performance would change. This not only deepened my technical skills but also gave me insight into the importance of data selection and how it influences the accuracy and relevance of machine learning applications. Overall, the combination of theoretical learning and practical experimentation made this a highly productive and engaging week.