AI is Rising - Buying Hardware Smartly? General - HWzone Forums
adplus-dvertising
Skip to content
  • Create an account
  • About Us

    Hello Guest!

     
    Please note - in order to participate in our community, comment and open new discussions, you must join as a registered member.

    Our members enjoy many advantages, including the ability to participate in discussions, enjoy raffles and promotions for members of the site, and receive our weekly content directly by email.

    Do not like being harassed by email? You can register for the site but do not submit your registration to the weekly email updates.

AI is Rising - Buying Hardware Smartly?


furyanaor
 Share

Recommended Posts

"Say? Which computer to buy? !!?"
The question is perhaps the most common here in the forum and in a few other places I came across ..
The thing is that the answer to this question is not just about finding the right computer, those focused on customer needs and the customer never knows what he needs!
So, we are in a world where AI is Rising and I have already come across two interesting solutions for laptops. One of the US retail giant Walmart:
Great interface and exceptional user experience, but the stock is subject to supply in the store, so the level of accuracy for me is lacking.
And the other is Choosist, a Beta startup company:
In my humble opinion, the customer journey is too long and tedious, the offer is subject to Amazon only and the prices are not up to date, so that's enough.
In short, as the guys who understand - try for yourself and tell me:
A. What is your impression of the solution to the question of characterizing the needs of use?
B. Do you see this as a successful substitute for consulting with a human specialist?
third. What do you think is the chance that non-technologists have the ability to pick out On their own will Mazap really use such a system or a similar application?
 
image.png.21125b699a8af91b2fe45eb25e7de3c1.png
Link to content
Share on other sites

First of all thank you for the references :)

 

In my opinion their performance is not the best in the world because the most important parameter here is not taken seriously enough and I am talking about the quality of accuracy.

But the question was on a more philosophical level - what do you think about the idea of ​​a machine performing an in-depth analysis

Draw a quality data-based conclusion in a way that is not biased by the opinion of one person.

After all, sometime - we will get there and I smell that this day is not so far away ..

 

And regarding AI Let me remove doubt that this is not just a few If Statements 

These in the ML Algorithm after the Data measurement phase from the User (which actually amounts to several Ifs)

A proximity test is performed to the nearest result of many species found in DATASET (bet on KNN)
Which by the way is a process very similar to what we do in our head when we think of a computer that fits the customer's needs

After all, we have the ability to meet the definition of Even though she's a bit lame .. hope I was able to convince 🙃

Link to content
Share on other sites

Quote of furyanaor

And regarding AI Let me remove doubt that this is not just a few If Statements 

These in the ML Algorithm after the Data measurement phase from the User (which actually amounts to several Ifs)

A proximity test is performed to the nearest result of many species found in DATASET (bet on KNN)

I was not convinced, I have a hard time seeing how a significant dataset was built here so that a valuable ML could be made about it.

Yes there is a database of data, someone gave them weights quite manually and computes a proximity function according to the user's choices but such algorithms were used long before and regardless of AI.

 

To your main question, I see no fundamental reason that AI will not be able to do this well but have a hard time seeing where they are coming from a dataset that he can learn from.

Link to content
Share on other sites

useless buzzword in terms of functionality is usually a very helpful component to PR
And when trying to penetrate the market with an innovative system then PR is a top priority
Never mind that ???? And regardless I have already heard quite a few experts who define every ML most simply as AI

 

As for the size of the dataset, it is quite huge when it embodies the entire range of laptops on the market

Of course with all the variants that the Vendor offers (once with 8GB of RAM and once 16GB and another combination with an I7 processor instead of I5)

When all this is sucked from some API with a petty database or even a crawler that runs on as the sites of each of the manufacturers

 

 

And as for the ranking (weights mentioned) then I did not aim for a manual ranking

Because it will include the distortion according to the opinion of one expert or another who will build it

These are based on the API of benchmarks such as userbenchmark

And from there derive a ‘pure’ result that I tend to believe will lead to a high and up-to-date level of accuracy

Link to content
Share on other sites

Quote of furyanaor

As for the size of the dataset, it is quite huge when it embodies the entire range of laptops on the market

It's not really helpful, the ML needs to learn the relationship between the input that is the questionnaire with the consumer's answers and the output that is what that person bought. Your dataset does not include neither this nor that and certainly it is impossible to learn from it the connection between the two.

Link to content
Share on other sites

Quote of etal

It's not really helpful, the ML needs to learn the relationship between the input that is the questionnaire with the consumer's answers and the output that is what that person bought. Your dataset does not include neither this nor that and certainly it is impossible to learn from it the connection between the two.

Why should he include what that person bought? The goal is that it's at the end of the day "knowledge" better than the average person isn't it?

In fact, the connection to synthetic tests is correct here in my opinion, but I'm very sure that this is not what was done there and in any case I do not need AI for it again (you can see the website of For example showing what FPS you will get with each machine they sell)

Edited By Moon-Mage
Link to content
Share on other sites

Quote of Moon-Mage

Why should he include what that person bought? The goal is that it's at the end of the day "knowledge" better than the average person isn't it?

One can take an expert who will give scores to those purchases and rate how much they really fit or even produce a collection of artificial samples, but in the end that’s how ML learns. How exactly do you propose to use ML in this case? 

Link to content
Share on other sites

Quote of etal

One can take an expert who will give scores to those purchases and rate how much they really fit or even produce a collection of artificial samples, but in the end that’s how ML learns. How exactly do you propose to use ML in this case? 

To give the desired results for each use and budget according to benchmarks, ie you need to have either access to such a pool of tests in the whole range of machines you want to sell or someone to do it manually and give up ML completely (much more logical).

 

Again I do not at all think there is room here to push AI of any kind, there are clear numerical results and there is really no need for learning here but for a really short work of one person who can do it in one day and enough (even if it means throwing what he reads on some site to excel table Simple).

Edited By Moon-Mage
Link to content
Share on other sites

Quote of furyanaor

From there derive a ‘pure’ result that I tend to believe will lead to a high and up-to-date level of accuracy

You are very naive about the ability of AI and about the results it gives, it learns the same biases that are in the dataset and depends entirely on how the human factor fed it the parameters. The company behind the algorithm, its goal is to maximize its revenue, not necessarily to give the user the most "accurate" result if such a thing exists at all.

Edited By etal
Link to content
Share on other sites

Quote of Moon-Mage

You should have or have access to such a pool of testing in the entire range of machines you want to sell

It is not wise to limit it to a very limited number of computers but want a system that will recommend all the variety sold in the market and there is no such exact comparison database.

Link to content
Share on other sites

Join the discussion

You can then join the discussion and then join our community. If you already have an account with us, please Log in now To comment under your username.
Note that: The comment will appear to the surfers after approval by the board management team.

guest
Add a comment

×   The content pasted is with formatting.   Remove formatting

  Only 75 emoji are allowed.

×   Your link has been automatically assimilated.   Show as regular link

×   Your previous content has been automatically restored.   Clear all

×   You can not paste images directly. Upload or insert images from URL.

 Share

  • Latest news

  • Buzz-Zone: Everything hot on the net

  • Popular now

×
  • Create new ...

At the top of the news:

Gaming treats for the weekend

Another free game from the Epic Games Store, free titles to try out from Steam and Ubisoft Connect, fresh game packages on the Humble Bundle website and more are waiting for you inside

new on the site