The First Madeira International Workshop in Machine Learning was held in the last week of July in the University of Madeira. The four days workshop was organized by IKnowD, a non-profit organization that helps scientific and societal development, through spreading research results.
What did we learn about Machine Learning?
What are the ANNs?
During the four days, we received an introduction to the basics of Machine Learning.
On the first day, we studied the Python coding language and the importance of artificial neural networks (ANN). The ANN model is a biologically inspired model, mimicking animals’ learning mechanisms by observing data.
A typical ANN is composed of an input layer, a hidden layer, and an output layer. To make the ANN work, first, the input layer feeds the hidden layer with data from the network. Then the hidden layer processes the raw information and sends the result to the output layer.
ANNs with only one hidden layer are called shallow networks, while a deep network has numerous hidden layers.
When the hidden layer receives the information, it arrives in different weights. The ANN adjusts to these weights, regarding the examples which were provided and the feedback from the errors. This is what we call training.
How did we use Machine Learning?
In the next three days, we checked how we can use Machine Learning in practice. The first example was image analysis by DeepFFNN (feedforward neural network). Then we proceeded with convolutional neural networks (CNNs) and long short-term memory (LSTMs) networks.
What was the outcome of the Machine Learning Workshop?
On our last day, we were ready to use our freshly earned knowledge in the form of a mobile app. The app was able to process the information of the picture of our pets and decide if we can see a dog or a cat.
Overall, we found the workshop very useful and interesting. We are looking forward to implementing what we learned, to make our products and services even better.
Floating Particle is a gold sponsor of the first Madeira International Workshop in Machine Learning.