# Processing with AI ## Exploration : IA & Ethics Name: > Samuel Lermytte > Subject: > Detect dermatological problems using Computer Vision >[TOC] ## Design brief ### Biais If we don't source our dataset with enough rigor, the following biais might appear: >1. overfitting >2. non representative dataset >3. over/under represented skin issues We will ensure that our model is not biaised by: >1. Sourcing our data from different databases located in different countries/continent (for example ) >2. Making sure our data take into account various minorities and ethnicities >3. Making sure that a wide range of skin issues are present in our database. ### Overfitting We will make sure our model does not overfit by > Checking the accuracy of our model on other open sources databases with random and diversified people ### Misuse >We have to remind ourselves that our application could be misused by fraudulent skin doctors to get their patients to believe they have a skin issue that they actually do not have. ### Data leakage >We have decided that our training dataset will be fully open-sourced, but before we made sure that it does not provide intimate information that could be used/provided unlegitimately (eg: intimate skin pictures/ revenge porn) ### Hacking > If someone found a way to "cheat" our model and make it make any prediction that it want instead of the real one, the risk would be that we fail to detect people with a skin issue, in the same way we'd fail to detect people with coronavirus and that they would start contaminating others.