ModuleNotFoundError: No module named 'sklearn.neighbors.classification'

[YePG]

how to solve it?

[siabingtao]

I am facing the same issue too. Is anyone able to help? This happens when I am trying to run the s5_test.py script.

[siabingtao]

Hi, I managed to rectify it. I think that this happened because the installed version of scikit-learn was a newer one.

I first uninstalled scikit-learn:
pip uninstall scikit-learn
I then proceeded to install scikit-learn version 0.20.4:
pip install scikit-learn==0.20.4

[steve-95]

I have try pip install scikit-learn==0.20.4, and solved my problem ,thanks @siabingtao

[muyou111]

to install scikit-learn version 0.21.3 is OK

[TYang92677626]

I'm using scikit-learn1.4 version
Without downgrading the version, change sklearn.neighbors.classification in the trained_classifier.pickle of the training model to sklearn.neighbors._classification
Because scikit-learn has been underlined since version 0.22.x.
There are also several underlined ones:
Original: sklearn.tree.tree
Change to: sklearn.tree._classes (needs to be changed to classes here)
Original: sklearn.ensemble.forest
Change to: sklearn.ensemble._forest
Original: sklearn.preprocessing.label
Change to: sklearn.preprocessing._label
etc.

[steve-95]

thanks for your reply, I have get a mastor with the help of your work. good luck!

…

---Original--- From: ***@***.***> Date: Mon, Jan 29, 2024 11:18 AM To: ***@***.***>; Cc: ***@***.******@***.***>; Subject: Re: [felixchenfy/Realtime-Action-Recognition] ModuleNotFoundError: Nomodule named 'sklearn.neighbors.classification' (Issue #76) I'm using scikit-learn1.4 version Without downgrading the version, change sklearn.neighbors.classification in the trained_classifier.pickle of the training model to sklearn.neighbors._classification Because scikit-learn has been underlined since version 0.22.x. There are also several underlined ones: Original: sklearn.tree.tree Change to: sklearn.tree._classes (needs to be changed to classes here) Original: sklearn.ensemble.forest Change to: sklearn.ensemble._forest Original: sklearn.preprocessing.label Change to: sklearn.preprocessing._label etc. — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: ***@***.***>

[hieu2k2boss]

Training_classifier.pickle

how to open file :"Training_classifier.pickle". I got this error when running "s5_test.py" .
Error:

Traceback (most recent call last):
File "/content/drive/MyDrive/Realtime-Action-Recognition-master/src/s5_test.py", line 357, in
dict_id2label = multiperson_classifier.classify(
File "/content/drive/MyDrive/Realtime-Action-Recognition-master/src/s5_test.py", line 219, in classify
self.dict_id2clf[id] = self._create_classifier(id)
File "/content/drive/MyDrive/Realtime-Action-Recognition-master/src/s5_test.py", line 201, in
self._create_classifier = lambda human_id: ClassifierOnlineTest(
File "/content/drive/MyDrive/Realtime-Action-Recognition-master/src/../utils/lib_classifier.py", line 135, in init
self.model = pickle.load(f)
_pickle.UnpicklingError: invalid load key, '\xef'.

[xionzhen]

hello,I met a question about how to open file :"Training_classifier.pickle" and change the pickle.If I just open it in a directly way and change it to sklearn.neighbors._classification,will get the error of _pickle.UnpicklingError: invalid load key, '\xef'.

I'm using scikit-learn1.4 version Without downgrading the version, change sklearn.neighbors.classification in the trained_classifier.pickle of the training model to sklearn.neighbors._classification Because scikit-learn has been underlined since version 0.22.x. There are also several underlined ones: Original: sklearn.tree.tree Change to: sklearn.tree._classes (needs to be changed to classes here) Original: sklearn.ensemble.forest Change to: sklearn.ensemble._forest Original: sklearn.preprocessing.label Change to: sklearn.preprocessing._label etc.