Sandeep Bhutani
  • Member for 3 years, 3 months
  • Last seen this week
  • India
Products classification by name
Accepted answer
5 votes

I think, and have done similar problem too, that this problem can be solved in this way: 1. Generate NGrams 2. Create 1 hot encoding matrix 3. Pass to Naive Bayes or Random forest It would ...

View answer
How should I implement machine learning for multi-tenant website?
2 votes

Ideally you should have the historical data for your model to gain trust. However, since you don't have data you should atleast add the tenant/region feature to your model and train again. Also keep ...

View answer
Audio Analysis : Segment audio based on speaker recognition
2 votes

Initially I did using silence detection but later moved to pyAudioAnalsis which is better. Check "Speaker Diarization" section in Segmentation in pyAudioAnalysis

View answer
Load keras model in Java
Accepted answer
2 votes

Giving you some links which have worked: deploying-keras-deep-learning-models-with-java : this uses DeepLearning4J, ND4J which supposedly you have tried already. tensorflow-keras-java And same ...

View answer
AllenNLP installation issue - No matching distribution found for torchvision<0.9.0,>=0.8.1
1 votes

Finally it worked using this command: conda install torchvision==0.8.2 -c pytorch then: pip install allennlp==2.1.0 allennlp-models==2.1.0

View answer
Deciding which samples the model will (probably) classify incorrectly
1 votes

You can give following a try : Extract features of your working images. - take out weights of last-1 layer Detect these features importance. You can use Eli5 for feature importance. Extract the ...

View answer
Understanding Classifier performance on text data
1 votes

As the data is imbalanced and skewed towards few classes, that's why RF and Logistic results are biased causing high FP values, therefore high precision and low recall. SVC on the other hand might ...

View answer
How to deal with memory exceptions while processing large amount of data?
1 votes

I had also faced similar issues when my python script was processing my 12GB and 35GB files. It used take weeks for processing and many times failed. You need to check your mathematical operation, ...

View answer
Generate new sentences based on keywords
1 votes

This is area of NLG . You can use template based text generation techniques, wherein you have defined structure of output text and fill in required blank areas based on keywords. This technique is ...

View answer
What is the best way for synthetic data generation while maintaining privacy?
0 votes

We solved this problem by using NER. Using spacy or similar alternatives, the entities can be detected and replaced with xxx . This way identification of company names, currencies etc becomes ...

View answer
Language translation with convolutional neural network
0 votes

CNN and RNN have different architectures and are designed to solve different problems. Images have a lot of pixels and thereby lot of features. Reducing some features doesn't impact much on what ...

View answer
where to store embeddings for similarity search?
0 votes

Why don't you cluster similar embeddings and store then use hashing to search faster. Then you can store them anywhere, maybe in bigdata hdfs distributed system for faster retrieval or simply hashed ...

View answer
Choosing a right algorithm for template-based text generation
0 votes

You give this one a try - generation-caption-for-new-image . This has 2 parts one labeling the image and second generating text for those labeled images. Your case is similar to this, the difference ...

View answer
How to model a binary classification problem in a evolving environment
0 votes

This is an interesting problem. It is analogous to an example of court case trial - say a person is not guilty until he is proved guilty (your NO label) and will stay guilty when proved (your YES ...

View answer
Is it safe to use labels created from unsupervised model to train a supervised model using the same data?
0 votes

Please take care of stratification while sampling data for training. You should mention stratify = y in train_test_split . High ROC-AUC may be pointing to imbalance dataset - when-is-an-auc-score-...

View answer
Identify specify areas in the text
0 votes

If you are still looking for answer then below are some suggestions. Please keep in mind that none of these will be 100% accurate: Train a classifier like RandomForest etc, on different known ...

View answer
tf.trainable_variables() returns blank list outside model_fn
Accepted answer
0 votes

This problem is resolved... tensorflow.python.framework.errors_impl.NotFoundError: Key output_bias not found in checkpoint – Actually there were model.ckpt-10949 files.. and tensorflow was picking up ...

View answer
Predicting High-School test scores after a disciplinary action
-1 votes

If your data is correct then i see a factor of gender also. I hope that is taken into account and also take Disp_action as categorical variable (one hot encoding). Also, try to club pre and post ...

View answer