Applied ML Author and Practitioner. I tweet to help beginners easily embrace ML & DS. Research Scientist @AmazonAds Tweets are my own.
Seattle, Washington
Joined November 2009
- Tweets593
- Following587
- Followers351
- Likes706
Jojo John Moolayil retweeted
. @jojo62000 offers a deep dive to understand how no-code platforms aid in facilitating accelerated ML adoption. buff.ly/3IYgaXT
Jojo John Moolayil retweeted
Thank you @Techstars for this incredible opportunity. A big shout out to my the wonderful @tuktucare team and all our champions. More strength to community powered care.
Elated to share that @Tuktu has been selected by @Techstars for the Techstars Future of Longevity Accelerator 2023 in partnership with @pivotalventures a @melindagates company techstars.com/newsroom/longe…
Jojo John Moolayil retweeted
What's the proper role for no-code ML platforms in a company's journey? @jojo62000 used to be a skeptic, but his thinking has become more nuanced over time. buff.ly/3CK4K66
A new post after a long while -
“No-Code ML Platforms: Boon or Bane?” by Jojo John Moolayil
link.medium.com/VWtZwjT8xwb
#machinelearning #datascience #nocodeml
Has the relaunch of Twitter Blue been better?
I wonder what additional steps have been taken to circumvent impersonation?
Phone number verification can definitely be diluted with burner accounts.
@TwitterBlue
The journey of learning to solve ML problems are heavily focussed on model building and undermines data exploration and validation.
In reality , the quantified efforts are often the other way round.
Validation >> Data wrangling/exp >> Model building
@TDataScience #AI #ml
Every 5th tweet in my feed is an Ad! Is this an intentional shift to encourage folks towards twitter blue that promises lower ads?
If yes, how much better would that experience be; cuz the current is definitely not what one wants.
#twitterads @elonmusk
What areas in Twitter have started to crack? @elonmusk so far spoke exclusively about improvements, there ought to be imp features that are failing.
To an extent, tech companies are bloated, but it's unlikely that twitter can thrive with a fraction of its workforce!
Thoughts?
Data science folks, what are the best resources that helped you embrace the field as a newbie?
If there are many, select the most impactful one for you.
#DataScience #MachineLearning #artificalintelligence
0%Video courses/blogs
0%University Course/degree
0%Kaggle/competitions
0%1:1 coaching
0 votes • Final resultsLarge Foundational Models in AI have made strides in abstracting key learning blocks for beginners.
Today, an enthusiast exploring NLP/computer vision problems, will almost always leverage an off the shelf model; mostly abstracting the science underneath.
Slowly, we will have a generation of applied ML folks comfortable in not knowing the details of some abstractions underneath. Akin to the abstractions in tech today; developers in 2022 have less foundational things to worry compared to 2 decades ago.
Jojo John Moolayil retweeted
In the last 10 years, as the Data Science field matured, what product/change do you think has been most influential in lowering the barrier to entry for DS as a career?
#MachineLearning #DataScience #ai #careers – at Vancouver, British Columbia
0%Python/R based ecosystems
100%Kaggle/ML Blogs/Videos
0%Research Papers/Academia
0%Cloud ( Sagemaker/others)
1 votes • Final resultsIn the last 10 years, as the Data Science field matured, what product/change do you think has been most influential in lowering the barrier to entry for DS as a career?
#MachineLearning #DataScience #ai #careers – at Vancouver, British Columbia
0%Python/R based ecosystems
100%Kaggle/ML Blogs/Videos
0%Research Papers/Academia
0%Cloud ( Sagemaker/others)
1 votes • Final results Jojo John Moolayil retweeted
Akin to all domains, problem solving stays at the heart of data driven decision making.
For data enthusiasts, exploring problems with an ML agnostic mindset is key and thus working backwards from the problem and not tech forward.
#DataScience #AI #ML
Akin to all domains, problem solving stays at the heart of data driven decision making.
For data enthusiasts, exploring problems with an ML agnostic mindset is key and thus working backwards from the problem and not tech forward.
#DataScience #AI #ML
Jojo John Moolayil retweeted
Data Science folks, I wanted to ask, do you see any mainstream ‘no code solutions’ working fine for your real world use- cases.
Are you comfortable in replacing your existing data science stack with it for a real world complex use case?
#DataScience #nocode #ml #ai
0%Yes, very comfortable
100%Na, only for demos
5 votes • Final resultsData Science folks, I wanted to ask, do you see any mainstream ‘no code solutions’ working fine for your real world use- cases.
Are you comfortable in replacing your existing data science stack with it for a real world complex use case?
#DataScience #nocode #ml #ai
0%Yes, very comfortable
100%Na, only for demos
5 votes • Final resultsA typical ML experiment mandates validating ideas faster. High velocity experiments enables several ideas to be validated faster and thus prune experiments bound to fail and double down on the ones ones bound to succeed faster.
#MachineLearning #DataScientist
High velocity experiments result in a fly wheel effect for machine learning. More experiments result in more hypothesis being validated which increases the exposure to better experiments and better hypotheses.