from imblearn.over_sampling import SMOTE from imblearn.under_sampling import RandomUnderSampler from imblearn.pipeline import Pipeline # Define features and target X = df.drop('infected', axis=1) y = df['infected'] # Split the data into training and testing sets X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42) # Define the resampling strategy over = SMOTE(sampling_strategy=0.5) # Oversample the minority class to 50% of the majority class under = RandomUnderSampler(sampling_strategy=0.8) # Undersample the majority class to 80% of its original size ..read more

I'm using gluonts.torch implementation of DeepAR to forecast a future sale using search volume as a dependent/input/dynamic variable. my data look like this: Date Sale Search 2018-12-30 205 13 2019-01-06 245 19 2019-01-13 207 20 2019-01-20 221 21 2019-01-27 179 17 ... ... ... 2023-11-26 142 11 2023-12-03 183 13 2023-12-10 211 14 2023-12-17 236 14 2023-12-24 275 15 I did the following: import pandas as pd import matplotlib.pyplot as plt import numpy as np from gluonts.dataset.pandas import PandasDataset from gluonts.torch import DeepAREstimator from gluonts.evaluation impor ..read more

I am working with audio classification using OPENSMILE library. After preprocessing the audio data i am getting a 800x25 shaped data which is just for one file (each files is around 15 seconds long) for training this dataset and for portability i want to convert to a CSV file, i have 5 folders in total (data/category_0 ... data/category_4) each category have around 50 files insides them (.wav) when processing one of the file data/category_0/audio1.wav after processing using Opensmile i am getting a pd data frame of shape (800x25) how should I train a multiclass classification when the data is ..read more

Practical Exam: House sales Real Agents is a real estate company that focuses on selling houses. Real Agents sells a variety of types of house in one metropolitan area. Some houses sell slowly and sometimes require lowering the price in order to find a buyer. In order to stay competitive, Real Agents would like to optimize the listing prices of the houses it is trying to sell. They want to do this by predicting the sale price of a house given its characteristics. If they can predict the sale price in advance, they can decrease the time to sale. Data The dataset contains records of previous ho ..read more

I'm trying to convert existing depth-anything PyTorch model to CoreML format. I decided to use Google Colab and took the following note for inferencing depth-anything model. However, I meet some exception while trying to import it on iOS side. Here is my code snippet for converting: # Installing all needed extensions !pip install coremltools # ... import coremltools as ct import torch # Convert the PyTorch model to TorchScript traced_model = torch.jit.trace(depth_anything, torch.rand(1, 3, 518, 518)) # Convert the TorchScript model to CoreML model_coreml = ct.convert( traced_model ..read more

How to programmatically set alerts for failure in a W&B run? I'm trying to set up alerts in my W&B (Weights & Biases) project to notify me if a run fails. I've been testing several functions I thought would work based on my research, but none seem to be implemented in the W&B API. Here's the code snippet where I attempt to set up these notifications: import wandb mode = 'dryrun' run_name = 'my_run' num_batches = 50 path = '/data' name = 'experiment1' today = '2023-08-01' probabilities = [0.1, 0.9] batch_size = 32 data_mixture_name = 'mix1' debug = mode == 'dryrun' run = wand ..read more

could someone help me out with my Pytorch installation? My device currently uses Windows OS and an AMD GPU. However, the Pytorch installation does not support Windows OS with ROCm combination. Only when Linux OS is chosen will the ROCm option be available. Can I use CUDA toolkit in replacement of ROCm? Or do I somehow change my OS to Linux? Is there some way to by pass all of these and still be able to use Pytorch? Any advice will be greatly appreciated! I have tried looking for installation tutorials on youtube but they do not have the same OS and GPU combination as I do. (That is Windows OS ..read more

I am researching methods to make an attendance system, where the professor clicks few photos (2 to 3) and upload on app where automatic attendance is given for about 80 students. I am limited training Data, which is the biggest drawback and major issue we need to counter. I made a basic model that trains and marks attendance. I need help to improve it and all the steps, like - How is CNN helpful here ? How do I train it to work like a Reward Punishment system where I can manually tell it who the person is it wasn't able to recognize so it learns on the way. Any help, advice or suggestions. Thi ..read more

I am working on creating a PCA Index in R to understand how it works. I've used 'make up' data for this purpose. However, when I plot the scores of the first component, the results are not as expected. Specifically, the score plot shows that the index performs poorly in the first year and then rapidly improves over time, which is the opposite of what I intended. pca <- prcomp(data, scale = TRUE, center = TRUE) score <- scale(data) %*% pca$rotation[, 1] plot(score) I'm trying to figure out what I might be doing wrong or how I can interpret the results better. Any insights or suggestions ..read more

I'm trying to implement a simple UNet with different modalities. The issue is training is taking way too long and I'm constantly running into OOM errors when num_workers>0. def train_batch(self, source, target,extra): with profile(activities=[ProfilerActivity.CPU],profile_memory=True, record_shapes=True) as prof: source_tensors = [tensor.to(self.device) for tensor in source] target_tensors = [tensor.to(self.device) for tensor in target] if ("rgb" in self.config["modalities"] or "semantic" in self.config["modalities"]): extra_tensors = [tensor.to(devi ..read more