◆ MATLAB

Implement cocktail party algorithm SVD

This is real work, not a feature someone invented — it comes from real job ads and real questions people asked. Below are four ready AI prompts: get it done, make it easy for the next person to say yes to, work out the right move when you are stuck, and stop it coming back.

4prompts
See the prompt↓All MATLAB tasks

The same task, four prompts

today's deadline · the next reviewer · the stuck moment · the pattern
AExecute — do the immediate taskImplement the cocktail party algorithm using Singular Value Decomposition to separate the two…+
Implement the cocktail party algorithm using Singular Value Decomposition to separate the two audio signals in 'mixed_signals.mat'. Store the separated signals in 'separated_audio.mat'.
when the reply comes backPush once: ask it to sharpen the weakest part, and to say what it assumed. Helpful?
BImprove — make it easier to acceptBefore I present these separated audio signals to Professor Davies, can you implement the…+
Before I present these separated audio signals to Professor Davies, can you implement the cocktail party algorithm using SVD? Make sure it handles the two mixed signals in 'mixed_signals.mat', and include a quick visualization of the separated waveforms to confirm the separation looks reasonable before I save them.
when the reply comes backPush once: ask it to sharpen the weakest part, and to say what it assumed. Helpful?
CDecide — diagnose the stuck momentI just ran the SVD-based cocktail party algorithm on 'mixed_signals.mat', but when I listened…+
I ran the SVD-based cocktail party algorithm, but the separated signals still sound like a mix of both sources.
I just ran the SVD-based cocktail party algorithm on 'mixed_signals.mat', but when I listened to the outputs, both 'separated_signal_1' and 'separated_signal_2' still sound like they contain both original sources. I'm afraid I might have made a mistake in applying the SVD components or the mixing matrix. How can I diagnose if the SVD decomposition itself was incorrect or if my reconstruction logic is flawed, without having to re-record the original, unmixed audio?
when the reply comes backPush once: ask it to sharpen the weakest part, and to say what it assumed. Helpful?
DBecome — change the patternI frequently implement signal processing algorithms, and they often produce outputs that are…+
My signal processing implementations often fail subtly, making it hard to pinpoint the exact error source.
I frequently implement signal processing algorithms, and they often produce outputs that are 'almost right' but not quite, making it incredibly difficult to diagnose where the error lies – in the theoretical setup, the data preprocessing, or the implementation itself. This costs me days of debugging. What habit should I change to systematically verify each step of a complex algorithm like the cocktail party problem, so I can catch errors early and accurately?
when the reply comes backPush once: ask it to sharpen the weakest part, and to say what it assumed. Helpful?

Questions people actually ask

honest answers, no sign-up

Every task here was seen in the real world. Someone doing the job named it, a real job ad asked for it, or a lot of people asked about it online.

If nothing real showed a task, it is not on the page. That is the whole rule.

They are the same job approached four ways, because what you need depends on where you are.

Get it done today. Make it easy for the next person to say yes to. Work out the right move when you are stuck. Learn the pattern so the job stops coming back.

For most of these jobs it can carry the heavy thinking - draft it, sort it, check it, rehearse it with you.

It cannot sit in your chair, take the blame when a number is wrong, or notice what nobody wrote down. Let it do the first 80%. Keep the last 20% that is truly yours.

No. Copy any prompt and paste it into the AI you already use. No account, no score, no wall in the way.

Any of them. The prompts describe the work rather than naming a product, so they are not tied to one assistant.

That is also why they keep working when you switch.

Change it freely. Every prompt is a starting line, not a rule.

Put in your real numbers, your real names and your real deadline. The more you make it yours, the better the answer comes back.

The tasks come from real job ads, published job data and the questions people ask in public forums.

The steps come from MATLAB's own documentation, with practitioner sources for the traps the manual does not mention.

Push once. Ask it to sharpen the weakest part and to say what it assumed.

Most wrong answers come from a missing detail rather than a bad prompt - tell it the thing it could not know.