README.txt

READ ME!

To run the ML algorithm

prerequisits
anaconda3 installed
set up your enviroment such that the following are installed:
numpy
matplotlib
collections
nltk
langdetect
sklearn

ensure you are located in the directory or a parent directory of the ML_CW.ipynb
ensure that the directory assignment-comp3222-comp6246-mediaeval2015-dataset submitted with ML_CW.ipynb is in the same directory as it



use command
>>>jupyter notebook

this will open up jupyter notebook
traverse to the ML_CW.ipynb file if in a prarent directory

- open it

you can run a block by using CTRL + ENTER or SHIFT + ENTER (the latter will move to the next block)

- all the blocks until you get to the block that starts with #ITERATION 1
- ensure that all blocks before this are run these are the functions that will be called, inspect them as you like,

all blocks after this will start with #ITERATION a.b
these blocks call on the previously defined functions to get the results
the each block represents either the end result of an Iteration or part of an Iteration.
- run the blocks of which the iterations you want to check

blocks that run preprocessing will take roughtly 5~15mins maybe longer or shorter depending on hardware
there is a progression bar to keep track of how far through it is.

when fitting an algorithm it can take several minutes.

The final 2 blocks will be the feature presentation, run these after all the function blocks if you want to see the feature graphs.