#tabular

1 messages · Page 1 of 1 (latest)

dreamy pumice
outer gazelle
next sky
#

yesss it worked

outer gazelle
#

The saddest channel on the saddest app I’ve ever used. 😢😢😢😭😭😭

next sky
#

OMG 😰

hushed sparrow
blazing pebble
#

I am working with tabular data from an old prediction competition. Will I need to submit my notebook and will the results be ranked?

sick basin
blazing pebble
sick basin
blazing pebble
sick basin
solemn plover
#

if anyone wants to experiment https://arxiv.org/abs/2209.15421 shows some promise (tabular data augmentation)-- it can help with really unbalanced datasets (or at least it did in my very unscientific test of 2 datasets)

arXiv.org
TabDDPM: Modelling Tabular Data with Diffusion Models

Denoising diffusion probabilistic models are currently becoming the leading
paradigm of generative modeling for many important data modalities. Being the
most prevalent in the computer vision community, diffusion models have also
recently gained some attention in other domains, including speech, NLP, and
graph-like data. In this work, we investi...

blazing pebble
#

I have another question. The dataset is normalized. Do you have any references for dealing with outliers on normalized data?

wary ridge
dark kiln
#

Hello, anyone loves LightGBM here ?

sick basin
dark kiln
#

XGBoost is amazing too of course

sick basin
#

Just wanted to ask whats special in lightgbm which you prefer ?

dark kiln
#

Just to make it clear, I am not saying "lightgbm is better than xgboost". Otherwise, i find lightgbm faster, it uses Exclusive Feature Bundling, i read about it and it helps with sparse features.

sick basin
#

You are right , but in my experience xgboost works better with sparse features

#

Also i should add , based on “No Free Lunch” theorem, the choice between XGBoost and LightGBM depends on the specific problem and the available data.

dark kiln
#

I agree, it depends on the data and the problem. Sometimes, when data contains too many categorical featurss, CatBOOST would be a good alternative.

hushed sparrow
#

So the question here was who loves LightGBM and we ended up with a conclusion that everyone loves at least some kind of GBMs?

sick basin
dark kiln
#

The conclusion that: Tabular ---> GBM

dreamy pumice
#

it seems to provide a good overview of the state of gbdts for tabular data

terse gale
dreamy pumice
true phoenix
#

So for tabular data gradient boosting is the way to go? 👀

dreamy pumice
#

Yes

silent sand
#

HI, I am Abdullah I am an ML engineer want to join any team to particapte in kaggle competions

true yoke
worldly pagoda
#

Has anyone come up with a method for sensitivity analysis that isn't too overly dependent on the model itself?

I was thinking of maybe making a separate polynomial-searching model to fit and get the gradients of for features with respect to the target feature, in an attempt to find some kind of non-linear trends independent of the production deep learning model, as just observing the gradients doesn't tell me what I want to know, otherwise not sure what else is out there

brisk plinth
#

Job Title: Part-Time Senior AI/ML Engineer (Remote)

We are seeking a skilled and experienced Senior AI/ML Engineer to join our remote team on a part-time basis. The ideal candidate will have a strong technical background, excellent communication skills, and the ability to work independently in a fast-paced environment.

Requirements:
-Minimum of 7–10 years of professional software development experience

-Proven experience working effectively in a remote environment

-Advanced English proficiency (C1 or higher); an American accent is preferred

-Availability to work 10–15 hours per week during EST or CST business hours

If you're a highly motivated engineer with a passion for building high-quality software and can commit to a flexible part-time schedule, we’d love to hear from you.
You can connect with me on WhatsApp: +1 (567) 469-5384

jovial kernel
#

Hi, @everybody
I have one question, I'm training ml models for the prediction, which is classification problem of 3 classes, where the number of samples are similar but the predition is skewed.
First class and second class is predicted with low precision tough, third class is never predicted. What's the reason? I can' t find the reason.
Before, when I applyed reinforcement learning, where the three classes were assigned to three actions and one action is never selected, too.
Actually, that is the preeiction model of forex eur/usd.

jovial kernel
#

I'm finding a US developer for the collaboration. If anybody interested, please dm me.

regal echo