Hackathon between Ubiqum Amsterdam/Berlin/Barcelona/Madrid to detect poisonous mushrooms from variables that can be detected from Image classification
- Goal - Minimize false positives (poisonous mushrooms that are classified as edible)
- Random forest used to get the most accurate results
- Logistic regression (manually tuning highly probable posionous mushrooms to poisonous)
- To exclude the possibility of any false positive