First of all,
I really appreciate your time and your dedication for the forum.
I really appreciate it.
As a practitioner,
I just want to sharing my understanding with you about your topic.
First,
the simpler, the better.
You have spent a lot of time to display tremendous criteria.
But the relationship between those variables hasn't been analyzed effectively.
Thus, your model couldn't tell us which criterion is the most influential decision variable.
Not to mention to give valuable forecasting information.
Second,
even with your assumption, you still has very limited legal space to collect relevant confidential information. Not to mention to make series comparison later.
The last but not the least,
Your topic is a dynamic, interactive process.
But your model is more likely than not stationary snapshot.
Without due diligence, your client normally would be overwhelmed by piles of information.
The challenging involved your project,
1. the utility of different variables.
e.g. some boys really care about girl's appearance, others don't.
meaning, some critical variables are not necessary in other candidates' model.
You never know.
2. Candidate changes. (dependent variable volatile)
Boys are changing. Their mindset changes. Their experience changes.
Their demand and vision change all the time.
It is hard to model it easily.
3. Information cost prohibitive
Rubbish in rubbish out.
In real world, it is hard to collect high quality information.
The time, the money and your energy.
All in all,
the beauty of model is its simplicity and its effectiveness.
You can definitely make it, but the price may keep you take conservative position before doing so.
Thank you buddy,
what's your opinion?
扫码加好友，拉您进群
全部版块
我的主页
转发别人看看！
