Infer information from Tweets. Useful for human-centered computing tasks, such as sentiment analysis, location prediction, authorship profiling and more!
Having a visual plot of performance vs. number of training examples used would aid the evaluation of a classifier or the feature extractor.
Ideally, the plot would be updated during the training process. That way, the researcher can see at a glance how the classifier is doing, and potentially decide on whether to cut training short or to let it continue.
Having a visual plot of performance vs. number of training examples used would aid the evaluation of a classifier or the feature extractor.
Ideally, the plot would be updated during the training process. That way, the researcher can see at a glance how the classifier is doing, and potentially decide on whether to cut training short or to let it continue.