Skip to main content

what i learned from Machine Learning Hackathon?

https://cdn-images-1.medium.com/max/1600/1*dE4F_nf8P60V2baaaOxgLQ.jpeg

First i will tell you what my topic is all about, my idea is to integrate AI with Restaurant based app

okay what is the big deal?
  • Using the bills we can predict how much raw material can be bought so we can minimize food wastage
  • We can also able to identify customer pattern which are more likely to buy after x item 
  • Last thing based on the feedback we can find mistake in us and improve it
Let's see what are the algorithm we are planned to use

Avoid Food Wastage: Using Time Series algorithm

For identify Customer pattern : FP Mining ( Frequent Pattern Mining ) is an algorithm that will predict the y after x items bought but we need to train it with existing data

Feedback : Sentimental analysis which separate the positive and negative comments then we will apply another algorithm to predict negative comments about which food item using naive bayes we give keywords such as Briyani,Rice.. the algorithm check whether it is available or not and how many times the keyword used

This what we planned,but we cannot complete the hack on time

Just explained our ideas to the people

Other cool project in the hackathon are
  • Chat Summarizer - Which learn to summarize the chat and present you with important message 
      • Usecase: when you have more than 60+ message most of them avoid reading it this project help them to read it 
  • Malware detection - This uses Navie bayes algorithm to learn malware detection 
  • Fill Form - You can fill a form just taking a picture of the data what you need to enter.Algorithm used - OCR (Optical Character Recognition) , address parsing , phone parsing 
  • Build a Form - Using OCR(Optical Character Recognition) you can able to construct a form 
  • BlackMirror - The project which is similar to TLDR in which you can paste the policy agreement it will summarize and tell you the essential thing in short and crisp manner
  • Markovian Chains for In-product Help
and lot more that's all i can remember

Lot of fun and cool experience

Attend lot of Hackathon find new friends and learn new things

Comments

Post a Comment

Popular posts from this blog

C program jackpot

/*Program to show sum of 10 elements of array & show the average.*/ #include<stdio.h> int main () { int a[ 10 ],i,sum = 0 ; float av; printf( "enter elements of an aaray: " ); for (i = 0 ;i < 10 ;i ++ ) scanf( "%d" , & a[i]); for (i = 0 ;i < 10 ;i ++ ) sum = sum + a[i]; printf( "sum=%d" ,sum); av = sum / 10 ; printf( "average=%.2f" ,av); return 0 ; } Output: enter elements of an array : 4 4 4 4 4 4 4 4 4 4 sum = 40 average = 4.00 /*Program to find the maximum no. in an array.*/ #include<stdio.h> void main () { int a[ 5 ],max,i; printf( "enter element for the array: " ); for (i = 0 ;i < 5 ;i ++ ) scanf( "%d" , & a[i]); max = a[ 0 ]; for (i = 1 ;i < 5 ;i ++ ) { if (max < a[i]) max = a[i]; } printf( "maximum no= %d" ,max); } Output: enter elements for array : 5 4 7 1 2 maximum no = 7 /*Swapp
Post a Comment
Read more

Newsboat RSS Reader

Let me tell what is rss reader then i will explain about newsboat,. RSS stands for Rich Site Summary. Simply it is way to subscribe to webpages and when new article is published you can see through the feeds. you get the idea Newsboat is a terminal rss reader which is simple, easy to use and highly customizable how to find rss feed link ?                                   simple trick that i used to find rss feeds in a website check for rss image like above [OR] Right click the website -> view page souce and search for rss copy the link and paste it in urls file in the .config/newsboat/urls (for linux and mac) RSS feeds from twitter: https://twitrss.me/twitter_user_to_rss/?user=<USERNAME> https://www.twitrss.me/twitter_search_to_rss/?term=<SEARCH TERM> example: https://twitrss.me/twitter_user_to_rss/?user=Hytale https://www.twitrss.me/twitter_search_to_rss/?term=Hytale RSS feeds from youtube: https://www.youtube.com/feeds/videos.xml?channel_id
Post a Comment
Read more

Ecosia

ECOSIA - A Green search engine “ Ecosia donates 80% profits to planting trees” Site: www.ecosia.org Lauched on 7 th December 2009 and created by Christian Kroll available in more than 26 language. It is located in Berlin, Germany.It is also called as CO2-neutral company.Ecosia has donated to different tree-planting programs. Until December 2010 Ecosia’s donations went to a program by WWF Germany that protected the Juruena-National park in the Amazonas. In order to make sure the protection was kept up, the program also drew up and financed plans with timber companies and the  local communities. According to B-labs, as of January 2015, "In donating 80 percent of its ad revenue, the search engine has raised over $1.5 million for rainforest protection since its founding in December 2009." According to Ecosia, by 2015, the search engine had almost 2.5 million active users, and searches through it had resulted in more than 2 million trees being planted.Since October 201
Post a Comment
Read more