## Projections onto a Convex Body

August 23, 2010

Given a point ${x} \in {{\bf R}^d}$ and an affine subspace ${H \neq {\bf R}^d}$, the projection of ${x}$ onto ${H}$ is the point ${P_H(x)}$ such that,

$\displaystyle \begin{array}{rcl} \vert \vert x - P_H(x) \vert \vert = \min \{ \vert \vert x y \vert \vert ~ \mid y ~ \in H \} \end{array} .$

One can obviously define other notions of projection but the above is probably the most commonly used in geometry. If ${x \notin H}$ then the projected point ${P_H(x)}$ is the unique point in ${H}$ such that the vector ${P_H(x) - x}$ is orthogonal to ${H}$. Also is well known that projecting any segment to an affine subspace can only shrink its length. The proofs of these facts are easy to see. But in fact these facts are just corollaries of the following two results.

 Given any nonempty closed convex set ${C \subseteq {\bf R}^d}$ and a point ${x \in {\bf R}^d}$, there is a unique point ${P_C(x) \in C}$ such that, $\displaystyle \begin{array}{rcl} \vert \vert x - P_C(x) \vert \vert = \min \{ \vert \vert x - y \vert \vert ~ \mid ~ y \in C\} \end{array} .$

and,

 Let ${C}$ be a nonempty closed convex set. The mapping ${P_C : {\bf R}^d \rightarrow C}$ is a contraction i.e. for any ${x,y \in {\bf R}^d}$ we have, $\displaystyle \begin{array}{rcl} \vert \vert P_C(x) - P_C(y) \vert \vert \leq \vert \vert x - y \vert \vert \end{array}$

Applying these results to affine spaces (which are nonempty closed convex sets) yields the results mentioned earlier. This projection that maps $x$ to $P_C(x)$ is known as the metric projection. The proofs of these facts are in order. Read the rest of this entry »

## Bounding the Volume of Hamming Balls

August 13, 2010

In this post I will derive an oft used inequality in Theoretical computer science, using the Chernoff technique. The following function turns up in a few places in combinatorial geometry and other places.

$\displaystyle \begin{array}{rcl} \Phi_d(m) = \sum_{i=0}^d {m \choose i}. \end{array}$

Here we assume ${m \geq d}$ since otherwise the function is trivially seen to be equal to ${2^m}$. Here we will show that,

$\displaystyle \begin{array}{rcl} \Phi_d(m) \leq \left ( \frac{me}{d} \right )^d \end{array}$

Two easy interpretations of ${\Phi_d(m)}$ are as follows. First, it counts the number of points in the hamming cube ${{\bf H}^m = \{0,1\}^m}$ whose hamming distance from the origin is at most ${d}$. This interpretation also allows us to derive the nice identity,

$\displaystyle \begin{array}{rcl} \Phi_d(m) = \sum_{h=0}^d \sum_{l = 0}^h {s \choose l}{m - s \choose h - l}. \end{array}$

for any ${0 \leq s \leq m}$. In the above expression we obey the convention ${{p \choose q} = 0}$ when ${p < q}$. To see this, interpret the above sum as the distance from a fixed point ${p \in {\bf H}^m}$ whose distance from the origin is ${s}$. The ${h}$ is the distance from ${p}$ and the inner summation counts the number of points at distance ${h}$. Then noting that the hamming cube looks the same from every point so the number of points of distance at most ${d}$ is the same from the origin or from ${p}$. Hence the identity.

This definition has a nice geometric interpretation because ${\Phi_d(m)}$ is also the number of points from ${{\bf H}^n}$ inside any ball of radius ${d}$. The second interpretation is as follows. Consider the discrete probability space ${\Omega = \{ 0, 1 \}}$ with the measure ${p(\{0\}) = p(\{1\}) = 1/2}$. Let ${X_1, X_2, \dots, X_m}$ be ${m}$ independent random variables ${X_i : \Omega \rightarrow \{0,1\}}$ defined as ${X_i(y) = y}$ for ${y \in \Omega}$. Let ${X = \sum X_i}$. Then we have,

$\displaystyle \Pr\{ X_1 = a_1, X_2 = a_2, \dots, X_m = a_m \} = \frac{1}{2^m},$

for any ${a_i \in \{0,1\}}$. As such we have,

$\displaystyle \frac{\Phi_d(m)}{2^m} = \Pr\{ X \leq d\}$.

We will use this interpretation to derive our inequality.

Lemma: ${\Phi_d(m) \leq \left ( \frac{me}{d} \right )^d}$.

Proof: Our first observation is that we may take ${d \leq m/2}$. The function ${\left ( \frac{me}{d} \right )^d}$ is an increasing function of ${d}$. Denoting this by ${f(d)}$ and extending it to a function on ${[0,m]}$ we have by logarithmic differentiation,

$\displaystyle \begin{array}{rcl} \frac{f'(d)}{f(d)} = \ln \left ( \frac{m}{d} \right ) \geq 0 \end{array}$

Thus for ${d > m/2}$ we have ${f(d) \geq f(m/2) \geq (2e)^{m/2} > 2^m}$. On the other hand ${\Phi_d(m) \leq 2^m}$ trivially.

Thus assuming ${d}$ is ${0 \leq d \leq m/2}$. Now we have the following identity,

$\displaystyle \begin{array}{rcl} \sum_{i=0}^d {m \choose i} = \sum_{j=m-d}^m {m \choose j} = 2^m \Pr\{ X \geq m-d \}. \end{array}$

so it sufficies to upper bound ${\Pr \{ X \geq m - d \}}$. We now use the Chernoff technique. For any ${\lambda > 0}$,

$\displaystyle \begin{array}{rcl} \Phi_d(m) &=& 2^m \Pr\{ X \geq m - d \} \\\\ &=& 2^m \Pr \{ e^{\lambda X} \geq e^{\lambda(m - d)} \} \\\\ &\leq& 2^m \frac{{\bf E}(e^{\lambda X})}{e^{\lambda(m - d)}} ~~ \text{[By Markov's Inequality]} \\\\ &=& 2^m \frac{\prod_{i=1}^m {\bf E}(e^{\lambda X_i})}{e^{\lambda(m-d)}} ~~ \text{[By Independence of the} ~ X_i] \end{array}$

Now for any ${i}$ we have,

$\displaystyle {{\bf E}(e^{\lambda X_i}) = \frac{e^{\lambda} + 1}{2}}$

and so we have,

$\displaystyle \begin{array}{rcl} \Phi_d(m) \leq \frac{(e^{\lambda} + 1)^m}{e^{\lambda(m-d)}} \end{array}$

We try to optimize the above expression by finding the ${\lambda}$ which minimizes the expression above. Noting that ${m/2 \geq d}$ we have, by differentiation and observing the sign of the derivative, that the ${\lambda}$ minimizing the expression is given by,

$\displaystyle \begin{array}{rcl} e^{\lambda} = \frac{m}{d} - 1 \end{array}$

and substituing that in the estimate we have,

$\displaystyle \begin{array}{rcl} \Phi_d(m) &\leq& \frac{\left ( \frac{m}{d} \right )^m}{\left ( \frac{m-d}{d}\right )^{m-d}} \\ &=& \left ( 1 + \frac{d}{m-d} \right )^d \left ( \frac{m}{d} \right )^d \leq \left ( \frac{me}{d} \right )^d. \end{array}$

The last step follows from the simple fact that,

$\displaystyle \begin{array}{rcl} \left ( 1 + \frac{d}{m-d} \right )^{m - d} \leq \left ( e^{\frac{d}{m-d}} \right )^{m-d} = e^d. \end{array}$

$\Box$

## Jung’s Theorem

August 7, 2010

Given a set of points of diameter ${D}$ in ${n}$ dimensional Euclidean space ${{\bf R}^n}$ it is trivial to see that it can be covered by a ball of radius ${D}$. But the following theorem by Jung improves the result by a factor of about ${\frac{1}{\sqrt{2}}}$, and is the best possible.
Theorem 1 [ Jung’s Theorem: ] Let ${S}$ be a set of points in ${{\bf R}^n}$ of diameter ${D}$. Then there is a ball of radius ${\sqrt{\frac{n}{2n + 2}} D}$ covering ${S}$.
Proof: We first prove this theorem for sets of points ${S}$ with ${\vert S \vert \leq n + 1}$ and then extend it to an arbitrary point set. If ${\vert S \vert \leq n + 1}$ then the smallest ball enclosing ${S}$ exists. We assume that its center is the origin. Denote its radius by ${r}$. Denote by ${S' \subseteq S}$ the subset of points such that ${||p|| = r}$ for ${p \in S'}$. It is easy to see that ${S'}$ is in fact non empty.

Observation: The origin must lie in the convex hull of ${S'}$. Assuming the contrary, there is a separating hyperplane ${H}$ such that ${S'}$ lies on one side and the origin lies on the other side of ${H}$ (strictly). By assumption, every point in ${S \setminus S'}$ is distance strictly less than ${r}$ from the origin. Move the center of the ball slightly from the origin, in a direction perpendicular to the hyperplane ${H}$ towards ${H}$ such that the distances from the origin to every point in ${S \setminus S'}$ remains less than ${r}$. However, now the distance to every point of ${S'}$ is decreased and so we will have a ball of radius strictly less than ${r}$ enclosing ${S}$ which is a contradiction to the minimality of ${r}$.

Let ${S' = \{ p_1, p_2,\dots, p_m\}}$ where ${m \leq n \leq d+1}$ and because the origin is in the convex hull of ${S'}$ so we have nonnegative ${\lambda_i}$ such that,

$\displaystyle \sum \lambda_i p_i = 0, \sum \lambda_i = 1$

Fix a ${k, 1 \leq k \leq m}$. Then we have,
$\displaystyle \begin{array}{rcl} 1 - \lambda_k &=& \sum_{i \neq k} \lambda_i \\ &\geq& \frac{1}{D^2}\sum_{i = 1}^m \lambda_i || p_i -p_k ||^2 \\ &=& \frac{1}{D^2}\left ( \sum_{i=1}^m \lambda_i (2r^2 - 2 \langle p_i, p_k \rangle) \right ) \\ &=& \frac{1}{D^2} \left ( 2r^2 - 2 \left \langle \sum_{i=1}^m \lambda_i p_i, p_k \right \rangle \right ) \\ &=& \frac{2r^2}{D^2} \end{array}$

Adding up the above inequalities for all values of ${k}$, we get
$\displaystyle \begin{array}{rcl} m - 1 \geq 2m r^2 / D^2 \end{array}$

Thus we get ${\frac{r^2}{D^2} \leq m - 1 / 2m \leq \frac{n}{2n + 2}}$ since ${m \leq n + 1}$ and the function $(x-1)/2x$ is monotonic. So we have immediately ${r \leq \sqrt{\frac{n}{2n+2}} D}$.
The remainder of the proof uses the beautiful theorem of Helly. So assume ${S}$ is any set of points of diameter ${D}$. With each point as center draw a ball of radius ${r = \sqrt{\frac{n}{2n+2}}D}$. Clearly any ${n + 1}$ of these balls intersect. This is true because the center of the smallest ball enclosing ${n+1}$ of the points is at most ${r}$ away from each of those points. So we have a collection of compact convex sets, any ${n+1}$ of which have a nonempty intersection. By Helly’s theorem all of them have an intersection. Any point of this intersection can be chosen to be the center of a ball of radius ${r}$ that will enclose all of ${S}$. $\Box$

## An IMO problem

July 20, 2010

I spent this weekend working on and off a very interesting IMO problem – problem 6 from IMO 2006. You can find that here . It was a real hard nut and only led to a crushed ego :(. Let me first describe the problem and a result I managed to prove in this direction, which was however much weaker than the problem statement.

IMO 2006 Problem 6 : Assign to each side $b$ of a convex polygon $P$ the maximum area of a triangle that has $b$ as a side and is contained in $P$. Show that the sum of the areas assigned to the sides of $P$ is at least twice the area of $P$.

I managed to prove the following result.

Theorem: Assign to each side $b$ of a convex polygon $P$ the maximum area of a triangle that has $b$ as a side and is contained in $P$. Then the sum of the areas assigned to the sides of $P$ is at least $\frac{3}{2}$ the area of $P$.

The proof is as follows. We prove first a small lemma.

Lemma 1 : Let $P$ be a convex polygon with $n \geq 4$ sides. Then there is a quadrilateral or a triangle with vertices among the vertices of $P$ that has area at least $\frac{1}{2}$ the area of $P$.

Proof: Let the polygon have vertices $A_1, A_2, \dots, A_n$. Without loss of generality assume that $A_1 A_k$ is the longest diagonal. The angles $\angle A_k A_1 A_2, \angle A_k A_1 A_n, \angle A_1 A_k A_{k-1}, \angle A_1 A_k A_{k+1}$ are all less than $\pi/2$ since $A_1 A_k$ is the longest diagonal. See the figure below for an example. For any vertex $A_i$ denote by $h(A_i)$ the distance from the line $A_1 A_k$. We will call the number $h(A_i)$ the height of $A_i$.

If $k = 2$ or $k = n$ then let $A_q$ be the vertex with maximum height. By our earlier observation regarding the angles at vertices $A_1, A_k$ with the line $A_1 A_k$, $P$ is enclosed inside the rectangle with base $A_1 A_k$ and height $h(A_q)$. As such the area of the triangle $A_1 A_k A_q$ is at least $\frac{1}{2}$ the area of $P$.

If $2 < k < n$ consider the vertex $A_p$ that has the maximum height among all vertices $A_i, 1 < i < k$ and the vertex $A_q$ that has maximum height among all vertices $A_i, k < i \leq n$. Clearly as before $P$ is enclosed inside the rectangle $R$ with a side parallel to $A_1 A_k$ through $A_p, A_q$ and the other sides through $A_1$ and $A_k$. Now

$\text{area}(A_1 A_p A_k A_q) = \text{area}(A_1 A_p A_k) + \text{area}(A_1 A_k A_q) \geq \frac{1}{2} \text{area}(R)$

We can now prove the theorem as follows.

Proof of Theorem: Let $A_1 A_p A_k A_q$ be a maximum area quadrilateral whose vertices are also vertices of $P$. We will call this quadrilateral $Q$. Clearly $\text{area}(Q) \geq \frac{1}{2} \text{area}(P)$ as we saw in the
previous lemma. The lemma actually showed that in few cases there
is actually a triangle $T$ with vertices among those of $P$ with area at least $\frac{1}{2}$ of $P$. We will not deal with the case of a triangle here since it is similar but we remark that if there is such a triangle, the theorem actually gets stronger and we can in fact prove what was required in the IMO problem.

Consider the triangles in the following four sets

$\left \{ A_1 A_2 A_q, A_2 A_3 A_q, \dots, A_{p-1} A_p A_q \right \}$

$\left \{ A_p A_{p+1} A_q, \dots, A_{k-1} A_k A_q \right \}$

$\left \{ A_k A_{k+1} A_p, \dots, A_{q-1} A_q A_p \right \}$

$\left \{ A_q A_{q+1} A_p, \dots, A_n A_1 A_p \right \}$

The sum of the areas of the triangles is at most the sum of the numbers on the edges $b$ of $P$. However, the sum of the areas is easily seen to be at least

$\text{area}(P) + \text{area}(Q) \geq \frac{3}{2} \text{area}(P)$

and we are done.

I found the solution to the IMO problem here and it is a real gem. It was a relief to see the solution!

## Some pointers for computer troublebombarding …

March 6, 2010

1) QTParted – free software for resizing partitions (it can handle NTFS partitions) found in Knoppix Free CD.
3) Broken exe associations – sometimes you may find all your shortcuts are behaving weird. You double click them and it asks you to indicate the program with which to open it. Well sometimes running them as Administrator works but this is a temporary solution. It may happen that even regedit or msconfig is not found when you type it at start! But dont panic – go directly to their location – search in C:\WINDOWS and right click on them and click “Run as Administrator”. So you can open the cmd prompt this way (CMD.EXE) and run a command like regedit or msconfig. Then to fix the broken exe associations you need to slightly tweak the registry. Read about it here – http://filext.com/faq/broken_exe_association.php.

It is good to have these handy – keep a copy of the softwares on a CD and a printout of the document. They can save hours of your time (and lot of sanity) sometimes.

## Illinois tollways

February 23, 2010

So you missed paying toll – you did not have enough cash or you just zoomed in through the IPass lane. If that happened on the Golden Gate Bridge , you have a toll violation and you will end up paying a hefty fine. I remember back in 2008 I missed paying the toll because I was like 10c short and I had to pay 25$. No checks, credit cards, invitation cards, greeting cards or poker face appeals worked. That really sucked! However Illinois has a great tollway system. Even if you speed past the toll booth for any number of reasons (including being in a hurry) you have upto 7 days to pay it online at http://www.illinoistollway.com/ – you just need to remember the approximate time, date and location of missed toll. There is a map on the website that one can use to figure out the missed booth you may need to spend 10 minutes on this but this is way more fair than paying 25$ fines! Go Illinois!

## Hello again!

February 6, 2010

It has been one and a half years since I made a post. The last post was on the tensor Calculus. While I have still not dropped plans to complete the series at some point of time, I think it is simply unsustainable for me right now. I have pondered once in a while at the reason why I stopped blogging and here are some possible reasons.

1. My aim was to make very expository posts. This is hard work and requires lot of motivation to keep up amidst all going on in life. Thus having this razor sharp focus on expository posts was not a good idea for my type of person. The moments of inspiration are few when I want to make such posts while a blog is something that should keep rolling.

2. I tend to plan quiet a lot before writing. This perfectionist attitude did not work well for me.

3. For sure I lack time management skills. But thats ok, I guess I am getting better.

I am not going to make any promises this time. Just try to write something whenever I feel like it … So once again – Hello everybody!