What you are going to read is a basic draft written in mostly one pass. (Disclaimer -- Some analogies may not be intuitively descriptive). If you know some terms related to 'Machine-Learnig' approach, then its an easy ride for you.. :)
[Note: in brackets you will find analogy to help easy understanding]
To go with Machine Learning Way:
"It takes time to
learn Classifying (understand) different
points (people). It takes a lot of
instances (experiences) to be able to build a good
classifier (impression). You can fasten the process by
Active-Learning (enquiring about someone) to learn quickly.
For many a
dataset (persons), the
initial-guess (First-Impression) decides how long will it take to get a perfect
classification (Faith). The
slope/gradient (small small acts) decides the
converging(goodwill) rate."
We apply Machine Learning, in a way Human thinks and percepts various things. Being human and our own Active Learner, why we take so much time to judge someone. We judge people, go along with them, but most of the time people later regret of having misjudged the person. Why is it so?? May be we are not seeing enough
features (taking kernels to higher feature space) of the person. Or may be
we (our classifiers) are not adaptive.
The classifier needs to be keep on trained, on not just
training set(inital impression), but also on the furthcoming
Test dataset (actual personality) so as to learn and adapt. The more time and instances we give to the classifier, the more we can understand the classifier.
The boundary values (or Support Vecotrs) are the good points which decide the classification hyperplane in Machine Learning. Similarly, its only in the time of crisis that u actually judge the real person. The character of the person is not what he does when everything is going right way. Its judged by how he handles pressure and behaves when things are not going right.
I always hope that people should not always form a rigid first-impression. Its important, but so is the time and reliability. Having judged a person in a long run (or having trained a good classifier), always will get you in touch with the right person. Be ready to accept them being a HUMAN. If Machine Learning can do it with a good accuracy, why dont we give chance to people around us.
[Please comment on the blog. Read it multi-pass and tell do u find analogy fine?]
Thanks,