Machine-Learning: The Alphabet

Machine-Learning: The Alphabet
My Inspiration
Some Sample Output

Inspiration

The meme above.

What it does

You, the sad, sorry user, input a word or multiple words, and the computer uses machine learning to tell you the next letter (alphabetically). Example: apple -> f, bottom -> n, jeans -> t.

How I built it

I used the far simpler solution to this "problem" (i.e. using a list of alpha characters and picking the next one) to create the dataset (10,000 random words from the nltk words corpora) that I trained the model with. Essentially, what one might do in 5 lines of code or less, I did in 120 lines, with machine learning.

Challenges I ran into

I forgot that input for models have to be the same size, so I had to pad my words to all be the length of the longest word (happened to be 23 chars in this dataset).

Accomplishments that I'm proud of

Having a functional machine learning model to tell you what you already know (the alphabet).

What I learned

Machine learning is dumb.

What's next for Machine-Learning the Alphabet

Put this algorithm in pre-school classrooms everywhere as a way to pressure children into learning the alphabet: "If a computer can do it, so can you, dammit!"