Multiple Solutions

After I established last week that I intended to employ supervised learning in order to realize my original work, this week, I decided to establish my fundamentals in supervised learning. I evaluated a research piece from the University of Pennsylvania that assessed artificial intelligence and machine learning as well as the nuances and relationships between the subsects of machine learning, supervised, unsupervised, deep, and reinforcement learning. As I extended my knowledge, it became clear that the researchers were asserting that practical applications with AI often involve multiple sub-architectures of AI. Thus, it is best to approach each application from a larger perspective and reduce focus on the individual architectures. In my case, my central focus needs to remain on how I am going to create a prediction model to read handwriting, not what type of machine learning I need to use. It is likely that I will have to use supervised learning to predict handwritten letters, input these results into a rudimentary neural network, and perhaps use unsupervised learning in order to discern the words from one another. Regardless of what I decide to use, these are simply blanket and abstract terms given to certain applications of machine learning, they will not dictate my approach to the problem. This will serve immensely valuable to the technical implementation of my project and I seem to be at the stage where I can officially put all the pieces together in an official program. The UPenn researchers concluded by tackling the larger implications of AI and the societal questions the technology poses. I really enjoyed the thought pieces the research provided as it conensided with my overarching intention to study artificial intelligence. How is AI going to fit into an increasingly competitive job market? How will bias impact machine learning and how can we democratize artificial intelligence? What will happen to humans as AI gets exponentially better? All of these questions are equally integral to the growth of AI as it trickles into daily lives and I am excited to evaluate my own answers to these dilemmas as I continue my work.