-
Noticed a minor problem in eval_knn.py with `--use_cuda False` for handing out of memory error.
It stops with the error:
```
File "eval_knn.py", line 164, in knn_classifier
retrieval_one_hot…
-
The original paper uses the following classifiers:
- C4.5
- 3-NN
- LOG
Unfortunately, the ML lib does not have those methods, so we will need to search for an alrenative.
Ideally, we want t…
-
When I trained on my customed dataset, the GPU memory allocated kept rising, even if I try to use `torch.cuda.empty_cache()`, this problem remained. And after I lowered the batch_size, the program was…
-
### Description
### 📌 Issues for Week 9 - Basic Projects of Machine Learning
🔴 **Project Title** : Pima Indian Diabetes Analysis
🔴 **Aim of the Project** : Creating a machine learning model, whic…
-
Description: Commonly in ML practice, it is not enough to run a single classifier. Through several research studies, it is shown that multiple classifiers working together can capture more complex kno…
-
Discovered a bug when trying to create a KNN model. KNN does not have a default constructor where Classifier.hpp tries to invoke the default constructor in k_fold_cv (which doesn't exist) and throws a…
-
When I ran Long Short Term Memory Classifier(Binary Classification), it showed Value Error. How can I solve this problem? Thanks Thanks so much!
--> 324 raise ValueError(f"Must pass 2-d inp…
-
Hello,
I am new to serializing models and trying to save a stacked classifier model.
```python
initial_types = [('feature_input', FloatTensorType([None, 8])),
('feature_inpu…
-
### 📌 Issues for Week 9 - Basic Projects of Machine Learning
🔴 **Project Title** : **Customer Segmentation**
🔴 **Aim of the Project** : Analysing different types of customers around the globe and …
-
Hello,
as we have already discussed on slack, it would be great if one could use all estimators and transformers in a sklearn pipeline e.g.
```
from sklearn.pipeline import make_pipeline
from …