NNC: Neural Network Classifier

• 

  Neural Network Classifier

Both myself (Kristiana) and Alexandra work with the public through extension. We work under researchers who are trying to change the world and make it a better place. As a wildlife ecologist we spend so much time looking through photos and videos for our desired data and the same is said for almost any agricultural research position. Most of us have little to know experience with coding and, even though we know that there is probably an easier way, it would take way to much time to learn what that is and how to do it.

So we decided to make a web app that makes machine learning more easily accessible to the average researcher. We created a web app where you input a folder with training photo's and it will teach the machine an algorithm designed for that person's unique data. This allows researchers to spend less time going through photo's and videos and more time doing what they love and changing the world.

We used node.js for the web app and python for the machine learning algorithm. Joseph used an example of a machine learning algorithm that categorized bees vs. ants as the base for the machine learning program. We used pug as the markdown language.

I have never coded before and have only been to one hackathon where I just collected data for my team. Alexandra has never coded before and is a first time hacker. So both of us had a steep learning curve trying to help out Joseph who has only been to 3 previous hackathons. We also had a hard time making such a broad algorithm to be able to accept and train completely different data sets. We had trouble executing python into node.js. We also got a call half way through the competition that my dog had gotten out and ran away so we had a hard time regrouping and focusing after that.

Not everything was bad though. Alexandra was able to learn a lot from the talks and expos. I was able to successfully create the front end with almost no experience in it. Finally, Joey was able finish the machine learning algorithm which is hard to do in such a short time.

Next for NNC, we plan on broadening it even more to be able to go through videos and to also group the photos instead of having to input one photo at a time.

Built With

• css
• dockerfile
• html
• javascript
• python