It feels great to be a part of Planet KDE! :)
I had submitted a proposal titled "Clone Tool for digiKam image editor" for Google Summer of Code (GSoC) 2011. Unfortunately, my proposal was not accepted. The results of GSoC and details of Season of KDE (SoK) 2011 came in together. I was still interested in working on digiKam so I planned to apply for SoK and spoke to some digiKam contributors. Marcel Wiesweg agreed to mentor me on a project - Face Recognition in digiKam.
I have been involved in some open-source projects only limited to use, evangelizing and bug reporting, but it will be my first time to actually contribute as a developer. In the coming months, I have lots of hacking to do. I shall post more on this in a while.
Sunday, May 15, 2011
Tuesday, December 28, 2010
Saturday, June 26, 2010
Summer of 2010
So the Summer of 2010 is almost over and it’s time for me to head back to college for the final year of my Bachelors. Time passes so soon...Here is a flashback of how I spent this summer break.
I have a passion for technology and wanted to take up some challenging research work in my area of interest this summer. I applied for the UGC Summer Internship Program 2010 at the Indian Institute of Science, Bangalore (IISc) under the Electrical Engineering Department and was among the 20 students in India to be selected for it. I worked under Prof. A.G. Ramakrishnan at the Machine Intelligence and Language Enginnering (MILE) Lab and was guided by one of his PhD. student, Thotreingam Kasar. My project was on Camera Based Document Analysis and Recognition, to be more specific, the topic was “Robust Text Detection and Extraction in Scenic Colour Images”.
I am very busy currently but would try to include the technical details of the project in the coming posts as I get time..
I have a passion for technology and wanted to take up some challenging research work in my area of interest this summer. I applied for the UGC Summer Internship Program 2010 at the Indian Institute of Science, Bangalore (IISc) under the Electrical Engineering Department and was among the 20 students in India to be selected for it. I worked under Prof. A.G. Ramakrishnan at the Machine Intelligence and Language Enginnering (MILE) Lab and was guided by one of his PhD. student, Thotreingam Kasar. My project was on Camera Based Document Analysis and Recognition, to be more specific, the topic was “Robust Text Detection and Extraction in Scenic Colour Images”.
I am very busy currently but would try to include the technical details of the project in the coming posts as I get time..
Wednesday, December 30, 2009
Why you should play the lottery
If you know anything at all about probability theory, you know that playing the lottery is a bad idea -- your chances of winning are very low. Or more precisely, if the lottery ticket costs $10, then your expected winnings are less than $10. Even so, I'm going to convince you that it is mathematically optimal for you to play the lottery.
Imagine that all you have is $1000 in cash. That's it. This is your whole livelihood. Would you bet it all on a coin flip? If you win, you get $2000; if you lose, you lose everything! Of course you wouldn't. You would rather spend that money on food for the next few days, while you are looking for a job.
Imagine instead that you are a millionaire, and you enjoy playing poker. Would you agree to play poker with your friends for $1000? Let's say they are of a similar skill level as you are, and they are fun people to hang around. Will you play? Sure. Maybe. Why not? It's only $1000. It's only 0.1% of your net worth. Your expected winnings are about $0 (breaking even), and the fun you will have playing is definitely worth more than that.
The point I am trying to make is that the first $1000 in your bank account is worth much more to you than your most recently added $1000. The value of money, to you, decreases the more money you have. Mathematicians call this "utility". If you buy something for $20, then you want that thing more than you want the $20. That thing has higher utility for you. The person selling you the thing wants the $20 more than he wants the thing. The couch has lower utility for him. Both sides win (increase their total utility).
Coming back to the money, it seems that the more money you have, the less is the utility of $1000 to you. In fact, things are a bit more non-linear. Imagine if you had $X in your bank account, for various values of X from $0 up to $1 million. What would your quality of life be? With $0, it would suck. Let's call that zero quality of life. With $1 million, life would be pretty sweet. Let's call that 100 units of quality. There is some non-linear, increasing function that maps money in the bank to quality of life. If you had nothing, and somebody gave you $X, the blue curve shows what your quality of life would become.
There are some bumps around $10,000 (the price of a car) and $100,000 (the price of a down payment on a house). This function is not concave.
Now imagine you already have $10,000 in your bank account. How much would your life improve if somebody gave you $10? Very little. $100? Still very little. $1,000? A bit. $1 million? It would go all the way up to 100 units. The red curve shows your new quality of life after receiving $X once you already have $10,000.
So what does this have to do with playing the lottery? Let's say a lottery ticket costs $10, and it gives you a small chance to win $1 million. That chance is almost certainly smaller than 1/100,000; otherwise, the lottery organizers have made a terrible mistake. Let's say the chance is half that -- 1/200,000. Should you play the lottery?
In terms of maximizing your wealth, no, you shouldn't. Your expected winnings are -$5 (negative $5). But what about quality of life? That's a lot more important than money! It depends on how much money you have. Let's say it is $10,000. What is your expected gain in quality of life? If you lose, you lose $10, which is worth almost nothing. If you win, you can buy a house. Your quality of life jumps all the way up to above 100. It sounds like you should play! In fact, it makes sense to play whenever you find yourself in one of the convex parts of the utility function that is curved enough to compensate for the negative expected value of the lottery.
In simple terms, if you don't mind risking $10 to win an amount of money that is larger than you would be able to earn at work, then sometimes it makes financial sense to play the lottery... :)
Imagine that all you have is $1000 in cash. That's it. This is your whole livelihood. Would you bet it all on a coin flip? If you win, you get $2000; if you lose, you lose everything! Of course you wouldn't. You would rather spend that money on food for the next few days, while you are looking for a job.
Imagine instead that you are a millionaire, and you enjoy playing poker. Would you agree to play poker with your friends for $1000? Let's say they are of a similar skill level as you are, and they are fun people to hang around. Will you play? Sure. Maybe. Why not? It's only $1000. It's only 0.1% of your net worth. Your expected winnings are about $0 (breaking even), and the fun you will have playing is definitely worth more than that.
The point I am trying to make is that the first $1000 in your bank account is worth much more to you than your most recently added $1000. The value of money, to you, decreases the more money you have. Mathematicians call this "utility". If you buy something for $20, then you want that thing more than you want the $20. That thing has higher utility for you. The person selling you the thing wants the $20 more than he wants the thing. The couch has lower utility for him. Both sides win (increase their total utility).
Coming back to the money, it seems that the more money you have, the less is the utility of $1000 to you. In fact, things are a bit more non-linear. Imagine if you had $X in your bank account, for various values of X from $0 up to $1 million. What would your quality of life be? With $0, it would suck. Let's call that zero quality of life. With $1 million, life would be pretty sweet. Let's call that 100 units of quality. There is some non-linear, increasing function that maps money in the bank to quality of life. If you had nothing, and somebody gave you $X, the blue curve shows what your quality of life would become.
There are some bumps around $10,000 (the price of a car) and $100,000 (the price of a down payment on a house). This function is not concave.Now imagine you already have $10,000 in your bank account. How much would your life improve if somebody gave you $10? Very little. $100? Still very little. $1,000? A bit. $1 million? It would go all the way up to 100 units. The red curve shows your new quality of life after receiving $X once you already have $10,000.
So what does this have to do with playing the lottery? Let's say a lottery ticket costs $10, and it gives you a small chance to win $1 million. That chance is almost certainly smaller than 1/100,000; otherwise, the lottery organizers have made a terrible mistake. Let's say the chance is half that -- 1/200,000. Should you play the lottery?
In terms of maximizing your wealth, no, you shouldn't. Your expected winnings are -$5 (negative $5). But what about quality of life? That's a lot more important than money! It depends on how much money you have. Let's say it is $10,000. What is your expected gain in quality of life? If you lose, you lose $10, which is worth almost nothing. If you win, you can buy a house. Your quality of life jumps all the way up to above 100. It sounds like you should play! In fact, it makes sense to play whenever you find yourself in one of the convex parts of the utility function that is curved enough to compensate for the negative expected value of the lottery.
In simple terms, if you don't mind risking $10 to win an amount of money that is larger than you would be able to earn at work, then sometimes it makes financial sense to play the lottery... :)
Monday, December 21, 2009
A test of common sense
Everybody has gut reactions. Have a look at the following two yes/no questions and tell me whether your first impulse is to say yes or no. I'm curious if anyone can "feel" the right answer to both questions, without having to think about it.
Question #1: I give you a fair 6-sided die. You inspect and weigh it and agree that it is fair. You roll it three times, and it shows the number 6 every time. I bet Rs.100 that it's going to be 6 again. Would you bet Rs.400 against me? (If it comes up 6, you lose Rs.400; if not, you win Rs.100.)
Question #2: Award-winning research has shown that 4 out of 5 math students at the University of California in Sunnydale wear white socks. We walk into a UC Sunnydale math class and look at the sock colour of 10 students. They are all white. I bet Rs.400 that the next student we check will also be wearing white socks. Would you bet Rs.100 against me? (If they are white, you lose Rs.100; if not, you win Rs.400.)
While we are in a betting mood, how much would you bet that your answers are correct? Probability is hard...
Post your answers as comments to this post and test your common sense.. :)
Edit: I have posted the answers to the questions as comments to this post..
Question #1: I give you a fair 6-sided die. You inspect and weigh it and agree that it is fair. You roll it three times, and it shows the number 6 every time. I bet Rs.100 that it's going to be 6 again. Would you bet Rs.400 against me? (If it comes up 6, you lose Rs.400; if not, you win Rs.100.)
Question #2: Award-winning research has shown that 4 out of 5 math students at the University of California in Sunnydale wear white socks. We walk into a UC Sunnydale math class and look at the sock colour of 10 students. They are all white. I bet Rs.400 that the next student we check will also be wearing white socks. Would you bet Rs.100 against me? (If they are white, you lose Rs.100; if not, you win Rs.400.)
While we are in a betting mood, how much would you bet that your answers are correct? Probability is hard...
Post your answers as comments to this post and test your common sense.. :)
Edit: I have posted the answers to the questions as comments to this post..
Wednesday, December 16, 2009
Cricinfo Rocks...
For all the cricket loving net savvy junta, Cricinfo is like God!! The wikipedia of cricket; first thing you would turn to if you are talking about cricket. Along with their live scores and awesome commentary, the thing which makes me a fan of Cricinfo is the quality of articles published on their site. Some of the articles published could be rated as classics in the true sense of the beautiful description, the elan, the charm which is present in the writing of the greatest writers. There are many articles worth saving (used to save them as Firefox bookmaks), but eventually lost them during Windows re-installation.
This is an abstract from one of the article which justified that Andrew Strauss is better placed than his predecessors to develop the winning culture his side so needs. The way the writer describes it is just awesome.. :)
"Durability counts among his (referring to Strauss) strong points, and it is an important quality in a captain. According to some, cricket captains merely walk out in smart blazers for the toss and are otherwise as powerless as traffic policemen in Kolkata. In fact, they take a hundred small decisions every day, must be able to make up their minds quickly and to live with the consequences. Before long, mental stamina becomes a factor. Happily, Strauss does not seem to be the tormented sort, forever agonizing over yesterday's mistake and reluctant to take tomorrow's decision. Nor does he appear to regard himself as a master tactician; merely as a common-sense leader able to absorb the blows and remain on track. There is no agitation in him. He may lose a match or his wicket. A proud competitor, he is not pleased by these turns of events but does not panic. It is not about him."
Another abstract of a fabulous article describing Sehwag's explosive batting as crimes against bowling humanity published recently..
"I speculated in my first World Cricket Podcast exactly what bowlers must feel when attempting to combat Sehwag on a good batting pitch. Suffice it to say that if this innings continues long into day three, the International Court of Human Rights may become involved, and the phenomenal Indian opener may find himself charged with crimes against bowling humanity.
For all the splendor Sehwag has once again given to the cricket-watching world, all record of this innings must be surreptitiously destroyed. What if impressionable young bowlers were to stumble upon evidence of the kind of abuse they may endure? What right-thinking parent would want their precious little baby bowler to grow up in such a heartless universe? Even bowling machines might refuse to bowl."
This is an abstract from one of the article which justified that Andrew Strauss is better placed than his predecessors to develop the winning culture his side so needs. The way the writer describes it is just awesome.. :)
"Durability counts among his (referring to Strauss) strong points, and it is an important quality in a captain. According to some, cricket captains merely walk out in smart blazers for the toss and are otherwise as powerless as traffic policemen in Kolkata. In fact, they take a hundred small decisions every day, must be able to make up their minds quickly and to live with the consequences. Before long, mental stamina becomes a factor. Happily, Strauss does not seem to be the tormented sort, forever agonizing over yesterday's mistake and reluctant to take tomorrow's decision. Nor does he appear to regard himself as a master tactician; merely as a common-sense leader able to absorb the blows and remain on track. There is no agitation in him. He may lose a match or his wicket. A proud competitor, he is not pleased by these turns of events but does not panic. It is not about him."
Another abstract of a fabulous article describing Sehwag's explosive batting as crimes against bowling humanity published recently..
"I speculated in my first World Cricket Podcast exactly what bowlers must feel when attempting to combat Sehwag on a good batting pitch. Suffice it to say that if this innings continues long into day three, the International Court of Human Rights may become involved, and the phenomenal Indian opener may find himself charged with crimes against bowling humanity.
For all the splendor Sehwag has once again given to the cricket-watching world, all record of this innings must be surreptitiously destroyed. What if impressionable young bowlers were to stumble upon evidence of the kind of abuse they may endure? What right-thinking parent would want their precious little baby bowler to grow up in such a heartless universe? Even bowling machines might refuse to bowl."
Friday, December 11, 2009
Subscribe to:
Posts (Atom)