ML Club Video: Word2Vec

Every ML model has something in common – numbers! Every weight, bias, neuron activation – from the inputs given to the final prediction – all them are represented as floating point numbers, often between 0 and 1. For quantitative data – like a patient’s heart rate and blood pressure – the problem is simple: just input the numbers into the model. For data like images, the data is represented as an array/tensor of the raw pixel values. But what about words? How do we represent words as numbers? Watch the video to find out!