Artificial Intelligence Against Homelessness and HIV

Artificial Intelligence (AI) can not only have its technical lead in the industry but also can help to improve one’s life. All it needs is a proper direction in to overcome research challenges and its thinking in a unified way.

AI fights for homelessness and HIV through the naive process. Initially, people talked with 4-5 homeless as their peers, but it wasn’t feasible because of the training cost. Bringing AI into the picture to capture which nodes influence the most leading to influence maximization problem, I strongly believe AI has the potential to reduce effects which could help us reduce homelessness. Bryan talked about the influence maximization algorithm which faces hardships due to poor records of no. of homeless which may require weeks of efforts to conduct in-person interviews. Bryan talked about how these AI techniques could then lead to community structure and friendship paradox. While community paradox has difficulty in a certain environment while the friendship paradox works by surveying a random node and then surveying its neighbor to generate a better result of influencing nodes. I think this could save our time and efforts and help us to put more effort in the areas where we are actually getting the results, in our case deciding upon the peer leaders.

I think there could be an alternative to this solution by combining tech with non-tech and believe government along with the educational institute could offer a helping hand to the homeless people. Having proper health education is at most important in this stage to prevent the spread of HIV. Instead of having peer leaders go to them, we could have weekend seminars for them in their locality and invite them over, collect their data and apply a machine learning algorithm to solve the issue in that area.

Moreover, collecting the family data from the city department we can establish a law to protect every family from homelessness by creating anti-homelessness service. The data collected could be passed through various machine learning algorithms including random forest, neural networks, boosted trees, and logistic regression. This would eventually help us to analyze that the families which are likely to face homelessness are also applying to anti-homelessness service. This would better define the goals of eradicating homelessness and HIV. The goal of this idea is to not only help the homeless but also to prevent homelessness.

I think Bryan did a great job explaining the present condition and various algorithms to tackle those conditions. The concepts were clearly articulated using algorithmic formula and graphs though I believe it could have got better if he had considered other cases as well, instead of just removing homelessness, prevention is equally important and necessary. His thoughts were well received by the audience, implementing such a great social problem using AI and Machine learning is a huge task. I really appreciate his work. Also, it would have helped him to demonstrate well if he had presented an interview with homeless people, it could have helped the audience to actually visualize the whole situation. Overall it was a great idea and I hope to see it as a big success.