Some security consultants working in the financial domain are currently advising a client who is investigating a potential money-laundering scheme. The investigation thus far has indicated that n suspicious transactions took place in recent days, each involving money transferred into a single account. Unfortunately, the sketchy nature of the evidence to date means that they dont know the identity of the account, the amounts of the transactions, or the exact times at which the transactions took place. What they do have is an approximate time-stamp for each transaction; the evidence indicates that transaction i took place at time ti ei, for some margin of error ei. (In other words, it took place sometime between ti ei and ti + ei.) Note that different transactions may have different margins of error. In the last day or so, theyve come across a bank account that (for other reasons we dont need to go into here) they suspect might be the one involved in the crime. There are n recent events involving the account, which took place at times x1, x2, . . . , xn. To see whether its plausible that this really is the account theyre looking for, theyre wondering whether its possible to associate each of the accounts n events with a distinct one of the n suspicious transactions in such a way that, if the account event at time xi is associated with the suspicious transaction that occurred approximately at time tj, then |tj xi| ej. (In other words, they want to know if the activity on the account lines up with the suspicious transactions to within the margin of error; the tricky part here is that they dont know which account event to associate with which suspicious transaction.) Give an efficient algorithm that takes the given data and decides whether such an association exists. If possible, you should make the running time be at most O(n2).