paul-tqh-nguyen reuters_topic_labelling issues

paul-tqh-nguyen / reuters_topic_labelling

Deep learning to automatically label Reuter's articles with the relevant topics.

1 stars 0 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Create condensed version of write up

#31 paul-tqh-nguyen closed 4 years ago
1
Refactor hyperparameter search to be able to search over all three architectures.

#30 paul-tqh-nguyen closed 4 years ago
1
Implement a dense layers neural network

#29 paul-tqh-nguyen closed 4 years ago
1
Investigate effectiveness of lemmatization

#28 paul-tqh-nguyen closed 4 years ago
1
Investigate approaches taken by others on multi-label problems.

#27 paul-tqh-nguyen closed 4 years ago
4
Investigate performance impact of replacing all numbers with a special token

#26 paul-tqh-nguyen closed 4 years ago
1
Investigate effectiveness of stop words.

#25 paul-tqh-nguyen closed 4 years ago
1
Refactor hyperparameter search to work with CNN functionality

#24 paul-tqh-nguyen closed 4 years ago
1
Abstract out a Classifier class

#23 paul-tqh-nguyen closed 4 years ago
1
Investigate CNN Archietectures

#22 paul-tqh-nguyen closed 4 years ago
2
Examine efficacy of older hyperparameter selection

#21 paul-tqh-nguyen closed 4 years ago
1
Optimize F1 Threshold.

#20 paul-tqh-nguyen closed 4 years ago
4
Evaluate Data Imbalance Impact

#19 paul-tqh-nguyen closed 4 years ago
4
Remove labels with insufficient data during pre-processing

#18 paul-tqh-nguyen closed 4 years ago
2
Determine <unk> words and evaluate their impact

#17 paul-tqh-nguyen closed 4 years ago
4
First pass at purely syntactically pre-processing the string data

#16 paul-tqh-nguyen closed 4 years ago
1
Implement feature to quit early upon convergence

#15 paul-tqh-nguyen closed 4 years ago
1
Investigate optimizing for the F1 score directly.

#14 paul-tqh-nguyen closed 4 years ago
2
Reimplement attention

#13 paul-tqh-nguyen closed 4 years ago
1
Create a hyperparameter search mechanism

#12 paul-tqh-nguyen closed 4 years ago
1
Track the best F1 score we've gotten so far.

#11 paul-tqh-nguyen closed 4 years ago
1
Use F1 score rather than accuracy

#10 paul-tqh-nguyen closed 4 years ago
1
Print the hyperparameters of our classifier prior to training process.

#9 paul-tqh-nguyen closed 4 years ago
1
Accomplish Write Up Goals

#8 paul-tqh-nguyen closed 4 years ago
1
Add functionality to load a model and evaluate texts

#7 paul-tqh-nguyen closed 4 years ago
2
Remove attention layers

#6 paul-tqh-nguyen closed 4 years ago
1
Do first pass at pre-processing data

#5 paul-tqh-nguyen closed 4 years ago
1
Create & deploy first pass at a document showing our findings.

#4 paul-tqh-nguyen closed 4 years ago
1
Create a classifier class to wrap our model

#3 paul-tqh-nguyen closed 4 years ago
1
Do first pass at updating the README

#2 paul-tqh-nguyen closed 4 years ago
1
Do first pass at model

#1 paul-tqh-nguyen closed 4 years ago
1