[12:28:48] While trying to understand and interpret the content of the output file of task 1 (part of outreachy application), I read about 'score', 'accuracy' and 'confidence' as related to a machine learning model. As I understand 'score' is the output value produced by a trained model on new input data. When we say, "the model is x% accurate", does that mean the prediction score falls within tolerable [12:28:54] error limits? [12:30:53] I find 'confidence' and 'accuracy' of an ML model are interwined concepts. They are subtle, but different in their own way. [12:33:29] I'll appreciate some help to understant these concepts. Can anyone share blog post, or a tutorial, or any resource that clearly explains 'score', 'accuracy' and 'confidence' terms as related to a ML model? [12:38:01] *error limits with x% chance [14:31:08] kalindi: for when you're reading again the logs. Welcome! It's nice to have you here. :) [17:01:48] pgadige: re your question about accuracy, score, and confidence. Can you copy/paste where you see the term confidence? because depending on the context, it may have a slightly different meaning. [17:02:50] pgadige: 'score' is, as you said, what your model predicts (the output of the model). For each data point (in the training and test sets), your model predicts an output and that is sometimes refered to as score. [17:04:19] pgadige: 'accuracy' is the fraction of predictions that the model gets right. (number of true positive and true negatives dividided by the total number of predictions). Check https://developers.google.com/machine-learning/crash-course/classification/accuracy [18:40:04] leila: I read about confidence and accuracy in a Q&A forum: https://stats.stackexchange.com/questions/310559/difference-between-confidence-and-accuracy [18:42:18] leila: Thank you for sharing that resource. I'm learning quite interesting things in the course, and having fun too by doing the examples. [19:52:23] pgadige: great to hear. :) [20:03:54] Hi, I am unable to view the messages here. Is there something that I am missing? [20:43:00] pandas_wiki: can you see this message? [21:05:54] yes this one [21:05:58] nice [21:06:25] pandas_wiki: can you explain what you mean by "I'm unable to view the messages here"? are you interested to read the previous conversations? [21:07:41] Hi is it possible to view the previous messages, I believe I am missing the important conversations [21:14:48] pandas_wiki: yes. all messages in this channel get logged and you can access the logs at https://wm-bot.wmflabs.org/logs/%23wikimedia-research/ (scroll down to the bottom of the page to see the latest) [21:15:23] pandas_wiki: if in the future you are looking for that link, check the "topic" of the channel. It's usually linked from there. (right now it is for example) [21:17:35] Ohh thanks alot :) [21:17:54] anytime. [21:40:14] I am facing some issues in running the github repository code [21:40:22] Can someone help me out? [23:29:54] pandas_wiki: the instructions for how to contact the mentors is provided in the task description in https://phabricator.wikimedia.org/T234519 .