Afiq Che Man
banner
mafqcm.bsky.social
Afiq Che Man
@mafqcm.bsky.social
🐍 Data science & machine learning. Studying 日本語. Random stuffs / games / anime off-work. Irregular reply sometimes. MPhil
MJIIT UTM
Reposted by Afiq Che Man
The US Army "is constructing an #AI and #ML ecosystem that will provide a trusted, safe environment where Army program managers and elements can bring in AI models and #test them against curated and trusted data." For more, see www.nationaldefensemagazine.org/articles/202...
#AI #DefenseAI
www.nationaldefensemagazine.org
January 13, 2025 at 7:13 AM
Reposted by Afiq Che Man
What #AI or #ML system would you use to train on #IoT data to maximise #smartfarm output? Asking for a friend
January 13, 2025 at 7:17 AM
smth i noticed online, there are generally two different camps on optimizing the modelling metrics (in my case, i was focusing on R2, MAE & RMSE).

group 1 - just go as low / high as possible bruh
group 2 - "it depends. look at the data first"
January 13, 2025 at 7:25 AM
Reposted by Afiq Che Man
sn-news: #nl #economy #ai #ml #llm Capturing the generative AI opportunity for the Dutch labor market
Capturing the generative AI opportunity for the Dutch labor market
Dutch businesses can embrace generative AI to speed up automation, increase productivity, and solve labor market tightness. What factors will affect adoption?
www.mckinsey.com
January 7, 2025 at 1:02 AM
Reposted by Afiq Che Man
Before Meta deleted its fake “AI” queer Black woman on Instagram I had a wild chat with it and uhhh, maybe it’s good actually..?
January 5, 2025 at 1:05 AM
Reposted by Afiq Che Man
We can pass a Polars DataFrame directly to XGBoost. This generally works fine with one important exception: categorical data. In a github issue an xgboost maintainer notes that with so many input types accepted it is a lot of overheard to support categorical data for each one!
January 6, 2025 at 10:30 AM
Reposted by Afiq Che Man
Every year at NeurIPS, I get a sense of where the community is headed. I'm so happy that the era of larger language models on larger datasets is coming to an end.
December 11, 2024 at 9:34 PM
It's (almost) year end season

Have you reviewed your models in prod yet?
November 22, 2024 at 3:49 AM
had to rework the script for model input preparation due to data being moved from sas to hive.
figured it out last night, but can’t run it till the evening bcs got stuck w other things.

finally got to it, went for a walk during the execution. came back to windows forced restart
November 20, 2024 at 11:11 AM
Reposted by Afiq Che Man
#firstpost
A #github repo which lists companies (or teams) that don't have broken hiring process.
#AI #AIjobs #ML #jobs
github.com/poteto/hirin...
GitHub - poteto/hiring-without-whiteboards: ⭐️ Companies that don't have a broken hiring process
⭐️ Companies that don't have a broken hiring process - poteto/hiring-without-whiteboards
github.com
November 19, 2024 at 11:59 AM
as usual, ran our model inference as part of BAU.
for this month in particular, the model inference kind of swapped in circumstances

the one that’s usually taking up so much time runs smoothly
while the one that’s usually running smoothly is now encountering a logging error 💀
November 18, 2024 at 6:55 AM
Reposted by Afiq Che Man
What are the main 𝗙𝗲𝗲𝗱𝗯𝗮𝗰𝗸 𝗟𝗼𝗼𝗽𝘀 𝗶𝗻 𝗠𝗮𝗰𝗵𝗶𝗻𝗲 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴 𝗦𝘆𝘀𝘁𝗲𝗺𝘀?

A regular Machine Learning Project consists of 4 main stages:

👇

#MLOps #AI #ML
November 14, 2024 at 12:47 PM
hey, i’m from X/Twitter. am a data scientist interested in AI research based in KL 🇲🇾
November 14, 2024 at 3:42 PM
Reposted by Afiq Che Man
Quite impressive to see the rise of users on Bluesky. I’ll post more often here. Maybe a place back for good ML/AI content?
November 13, 2024 at 7:48 PM