Difficulty in Exploratory Data analysis and Modeling for a newbie

dataexploration
sas

#1

HI,
I am an infant in Data analytics and started my analytics training from jigsaw academy in data science with sas, Financial analytics and Fraud analytics a week ago. I am already a certified base and advance sas programmer so the programming part is ok for me.

The problem is that I am finding it difficult to perform Exploratory Data analysis and I am not sure how to select proper variables for modeling process. I also have no domain experience. well its been only 2 days that I completed my Data preparation class and I created my first model today but I am finding my models to be bad.

I got an accuracy of 85% and was expecting some what 90-95%. I am spending 8-10 hours a day in studying analytics so that i can enter the analytics market quickly and with some good level of expertise. Looking at data for 48 hours and now I feel like I am super saturated. I see people here discussing super big topics and I cant create a simple linear regression model.

Does this happens to every one or am i the only one facing this. :expressionless:

suggestions please

regards
Neel


#2

@Neel_Biswas

Welcome to the club! I actually think that an 85% accuracy for a first model is quite good! Although I am not sure how do you define accuracy for a linear regression model.

As far as the experience you are undergoing, I think that every one faces this situation at some point. Some face it early in their career, some face it while learning more advanced tools or techniques. There is nothing to get dis-heartened about it. What looks difficult today provides you the learning and prepares you for tomorrow. Remember the first day of cycling or skating?

So, I would strong recommend that you keep your focus and efforts on, take help of these diiscussion portals and the Jigsaw ecosystem and I am sure you will come up to speed shortly.

@anon / @karthe1 - any tips you may want to share?

Kunal


#3

@kunal, I don’t have a lot to add beyond what you have already stated. Like for most things in life persistence (and interest!) is the name of the game.

To @Neel_Biswas, the self-confessed infant, I would say that you’re trying to to run even before learning how to walk. :wink: Although, I must say that for someone who’s just a week old, you’re doing all right. :smile: Take a break, do something else that lets your mind wander a bit, then come back and have a go at it.


#4

I second @kunal sir and @anon. I am almost with data sciences for the past 2 years (including one year into academics), but still every day in and day out is quite challenging. As both of them agreed, everyone would be facing the same level of difficulty unless you are hand-hold by an experienced personal . We cant expect any major support at this point of time, excepting the forums like AV, as the industry by itself is in nascent stage. Perhaps thatt is how any industry would shape up and perhaps that is how every technocrat jumping into any new area shall be facing the kind of nightmare that you and me facing.

Its good that you are putting in a_LOT_of effort. Keep this spirit up and see yourself as a leader-in-waiting in this space.

Good luck,


#5

thanku @karthe1 , @anon and specially @kunal for telling me to post it in the discussion section.actually i stopped for a day, rewatched a Game of throne episode and jst tried creating the model and i dont know from where i did it without any sweat.
thanx for the motivation people.
:smile: