Uff... Overfitting & Underfitting !???
Ruvva Poojitha
Intern at Kroll Ex-GDSC Data Science lead Student at ADITYA INSTITUTE OF TECHNOLOGY & MANAGEMENT, K.Kothuru, Tekkali - 532201, Srikakulam District, A.P(CC-A5)
You may think that this is just overfitting and underfitting right!! But, sometimes when the professor asks us about what we think soo simple, these simple fittings only confuse us like hell when needed. So let us explore with simple examples...
Overfitting
Overfitting is a scenario, where the model well trained on the training dataset, can perform well with the training dataset and perform very poorly with the test data set and unknown too...
Example : I got the maths question paper before the day of the examination, I prepared it very well, and the question paper was changed the next day,(i.e., the values only changed).. But i failed to answer the questions ,..uff??Thats because of overfitting, I practiced the questions very well and only can answer them, but if the value is changed a bit, I cant answer at allll??
Underfitting
Underfitting is a scenario where the model has less data to learn and can't perform well on the training data and also with the testing, unseen data also....
Example : During my language exam in my school, I met with a question in my test exam which I answered wrong. I should learn at least for my final exam right!! No!No! I cant do such good things,.. The same question repeated in my final exam and again i failed to answer this question ??
One more similar example,I failed my exam last year, and even the same question paper repeated this year, I failed again
I failed in training ??and also in testing ???? due to lack of learning ??
So, learn well guys, and don't overfit or underfit tooo....Learn what is required and know where to apply...????So you wont be failed in the exam like me?? All the best??
Note ????: Dont my story serious, but consider the quote of learning and learn wisely.????Happy Learning!!1