55,608 questions
- Bountied 0
- Unanswered
- Frequent
- Score
- Trending
- Week
- Month
- Unanswered (my tags)
Score of -1
0 answers
72 views
future_invoke_map() deprecated, cannot get replacement to work
purrr and furrr have deprecated some of the functions I previously used to build ensemble machine learning models, namely future_invoke_map() which I used as as noted in the snippet below, this mapped ...
- reputation score 9
Advice
0
votes
0
replies
56
views
Spent months building optimizers/CNNs from scratch in NumPy/CuPy — not sure what to build next, would appreciate direction
I have been teaching myself ML by building everything from raw math no heavy libraries like PyTorch, just NumPy/CuPy and derivatives worked out by hand. Wanted to share where I've landed and get some ...
- reputation score 1
Best practices
1
vote
2
replies
72
views
Why not use only non-missing value rows as the test set when evaluating an imputed ML pipeline?
For context, I am learning about imputation, and I want to understand what is wrong with my understanding.
The usual workflow I see is:
1. Split X and y into train/test, for example 80/20.
2. ...
- reputation score 1
Advice
0
votes
1
replies
63
views
Evaluating long-term memory limits in stateless LLM chatbots — feedback needed
I’m working on a research project exploring how stateless LLM-based chatbots handle long conversations and whether important earlier information is still reliably retained over time.
My idea is to:
...
- reputation score 1
Advice
0
votes
3
replies
62
views
class_weight vs data augmentation for handling class imbalance in binary classification?
I'm working on a face mask detection project using MobileNetV2 transfer learning for binary classification (with_mask vs without_mask). My dataset has a significant class imbalance:
With mask: 685 ...
- reputation score 1
Advice
0
votes
3
replies
115
views
How can I improve the accuracy of a Random Forest model for student performance prediction?
I am a beginner learning machine learning.
I built a Random Forest classifier to predict student performance using a dataset from Kaggle. My model currently achieves about 87% accuracy.
I would like ...
- reputation score 1
Score of -3
0 answers
85 views
How to efficiently stream sensor data from Arduino to Python for real-time AI analysis? [duplicate]
I am working on a project involving an Arduino microcontroller and a Python-based AI model. My goal is to use the Arduino to read sensor data and send it to a PC via serial communication (UART) for ...
- reputation score 1
Best practices
0
votes
2
replies
91
views
How to update dynamic user embeddings with negative ratings in 768-d space without causing vector drift?
I am building a production-grade recommendation system for a short-video platform (processing around 50k videos). The architecture utilizes a vector database (Qdrant) to store and query 768-...
- reputation score 1
Advice
1
vote
3
replies
89
views
Uncertainty Estimation vs Oversampling
I am currently doing some work with a fraud detection dataset as part of a research to leverage uncertainty to improve neural networks ensemble of experts' results.
Firstly I had to take the dataset's ...
- reputation score 1
Advice
0
votes
1
replies
74
views
Trigram Language Model :Two implementations give different loss, are they equivalent?
I am implementing a trigram character-level language model following Andrej Karpathy's makemore series. I have two implementations and I want to understand if they are mathematically equivalent or ...
- reputation score 1
Tooling
0
votes
0
replies
66
views
Best pre-trained vision model for multi-plant disease detection in async web back-end
I'm building a web app with FastAPI + async/await Python backend. Users upload leaf photos via API and the server should return: 1) plant species, 2) disease label or "healthy".
Constraints:
...
Score of -1
0 answers
110 views
Why is the cost of my neural network inconsistent (and sometimes increasing)?
I tried to follow this crash course to create a neural network from scratch. It seems to be working, which is great, but as I kept running the simulation I noticed that the cost of the network ...
- reputation score 43
Advice
0
votes
1
replies
92
views
Why is it so difficult to train an accuracte GAN model?
I'm trying to train a GAN model, but its results are very bad. The Generator doesn't seem to work. Can someone suggest how this can be improved? What is the best way to remember the code? Can this ...
- reputation score 1
Advice
1
vote
2
replies
101
views
Will a 80 GB GPU and a 48 GB GPU give identical results on an open source text-to-video model for the same quantization and seed?
I am considering to buy GPUs for my project of open source text-to-video models like ltx-2-19b (lightricks) or wan-v2.2-a14b. I read online that the same configuration/quantization and seed will give ...
- reputation score 49
Advice
1
vote
8
replies
173
views
What should I focus on?
I’m a 1st-year B.Tech CSE (AI & ML) student, and my first year has just ended. During this year, I mainly focused on learning Python because it’s important for AI/ML.
Now during my summer break, I’...
- reputation score 1