Subscribe to DSC Newsletter
Neeraj Sharma
  • Female
  • Bangalore,Karnataka
  • India
Share on Facebook
Share

Neeraj Sharma's Friends

  • Okan OYMAK

Neeraj Sharma's Discussions

Dealing with no data

Started Aug 17 0 Replies

I am working on predictive maintenance and get temperature data from assets. In few months or few days asset remains down and we do not get temperature value. In this scenario i cannot fill data with…Continue

Face Recogniser

Started this discussion. Last reply by Piyush Agarwal Feb 1. 1 Reply

I am developing face recogniser and in my data set there are 3000 different people each having around 40 different pictures. My question is, will my output layer have 3000 different classifications.Continue

Visualisation of train and test error

Started this discussion. Last reply by Lance Norskog Feb 8. 1 Reply

I am working on Digit Recogniser and after training the fully connected model with 3 CNN layers and 2 dense network layers with rely and softmax  activation functions, getting below curve.…Continue

Alternate of gradient descent

Started this discussion. Last reply by Neeraj Sharma Sep 16, 2018. 8 Replies

Hi Everyone,Recently i gave an interview for data science opening and interviewer asked optimisation algorithm for logistic regression. I answered gradient descent and explained same.He said these…Continue

Tags: algorithms., optimisation, regression, logistic, datascience

Gifts Received

Gift

Neeraj Sharma has not received any gifts yet

Give a Gift

 

Neeraj Sharma's Page

Latest Activity

Neeraj Sharma posted a discussion

Dealing with no data

I am working on predictive maintenance and get temperature data from assets. In few months or few days asset remains down and we do not get temperature value. In this scenario i cannot fill data with missing value techniques. Also cannot give some number because even 0 and -1 are valid values for temperature. How to deal with such data? I am thinking of putting very big value for such columns which is not possible as temperature. Please suggest.See More
Sunday
Lance Norskog replied to Neeraj Sharma's discussion Visualisation of train and test error
"It looks like your model becomes overtrained right around "3". That is where the evaluation curves diverge."
Feb 8
Piyush Agarwal replied to Neeraj Sharma's discussion Face Recogniser
"Ideally Yes, as you are trying to predict 3000 different classes (or people in your case). Your output layer should use "softmax" activation which will give you a probability of which class each input image belongs to. Take a look at this…"
Feb 1
Neeraj Sharma posted a discussion

Face Recogniser

I am developing face recogniser and in my data set there are 3000 different people each having around 40 different pictures. My question is, will my output layer have 3000 different classifications.See More
Jan 31
Giuseppe Longo liked Neeraj Sharma's discussion Visualisation of train and test error
Jan 22
Tim Matteson liked Neeraj Sharma's discussion Visualisation of train and test error
Jan 17
Neeraj Sharma posted a discussion

Visualisation of train and test error

I am working on Digit Recogniser and after training the fully connected model with 3 CNN layers and 2 dense network layers with rely and softmax  activation functions, getting below curve.Screen%20Shot%202019-01-17%20at%209.12.56%20PM.pngWhat i understood is model has high variance problem because after some time gap between training loss and validation loss is increasing.…See More
Jan 17
Neeraj Sharma replied to Neeraj Sharma's discussion Alternate of gradient descent
"That is a very nice article. Thanks for sharing."
Sep 16, 2018
Karthik PG replied to Neeraj Sharma's discussion Alternate of gradient descent
"I don't agree with the interviewer fully. ADAM, ADAGRAD,SGD are some of the popular ones used in solving real world problems when employing regression (of any kind) techniques...May be there is a specific problem he has in mind...."
Sep 4, 2018
Neeraj Sharma replied to Neeraj Sharma's discussion Alternate of gradient descent
"I also think so will try to compare both stochastic and swarm. Thanks"
Sep 2, 2018
Neeraj Sharma replied to Neeraj Sharma's discussion Alternate of gradient descent
"Thanks Vincent will try to implement and replace with gradient descent."
Sep 2, 2018
Eduardo Di Santi replied to Neeraj Sharma's discussion Alternate of gradient descent
"I think the interviewer its confused ... Stochastic gradient descent is heavily used and its the "de facto" method for learning Of course there are very good alternatives like Swarm optimization, but I will not say that nobody use SGD…"
Sep 2, 2018
surya prakash Sahu replied to Neeraj Sharma's discussion Alternate of gradient descent
"I think the interviewer was looking for some variation of gradient descent, because according to my knowledge gradient descent is used and used heavily; but with some tweaks. Check this momentum-rmsprop-and-adam, it'll give the idea."
Aug 31, 2018
Matt Wenham replied to Neeraj Sharma's discussion Alternate of gradient descent
"Genetic Algorithms may also be of interest. In general, the problem domain is that of Global Optimization."
Aug 29, 2018
Vincent Granville replied to Neeraj Sharma's discussion Alternate of gradient descent
"Here is a great alternative: swarm optimization, see https://www.datasciencecentral.com/profiles/blogs/swarm-optimization-goodbye-gradients The animated picture below shows how it works:"
Aug 29, 2018
Neeraj Sharma's discussion was featured

Alternate of gradient descent

Hi Everyone,Recently i gave an interview for data science opening and interviewer asked optimisation algorithm for logistic regression. I answered gradient descent and explained same.He said these days nobody use gradient descent? Is that true and if yes what are alternatives.What are disadvantages of gradient descent because of which it is not used anymore?Thanks for your help in advance.See More
Aug 28, 2018

Profile Information

Short Bio
Working till now on JAVA but very much passionate for data analysis. From last 8 months going through various courses for data science. Trying hard to work full time on data analysis.
My Web Site Or LinkedIn Profile
http://https://www.linkedin.com/in/neeraj-sharma-00788ba/
Professional Status
Technical
Years of Experience:
10
Your Company:
Cisco
Industry:
Software
Your Job Title:
SE III
How did you find out about DataScienceCentral?
via articles posted on Linked In
Interests:
Finding a new position, Networking, New venture
What is your Favorite Data Mining or Analytical Website?
http://www.datasciencecentral.com/,https://www.r-bloggers.com,https...
What Other Analytical Website do you Recommend?
http://https://www.datascienceweekly.org/

Comment Wall

You need to be a member of Data Science Central to add comments!

Join Data Science Central

  • No comments yet!
 
 
 

Videos

  • Add Videos
  • View All

© 2019   Data Science Central ®   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service