- In machine learning after splitting the dataset into train and test set can we directly import the required components like imputer dataframeselector and directly use the pipeline to compute the thing instead of doing it step by step…?
- In training the dataset we just segregate one column for the train set and we train the set using two datasets like why are we segregating into two datasets, like on passing the segregated whole dataset are we training our model to predict single column which we are passing along with it…??
- To completely train a model first separate it into test and train set, next just pass it through the pipeline, next train the model and now if we fine-tune is that all that we have trained a model…??
- After training the model sir has not completely explained about fine-tuning the model as I am a beginner to the programming world i didn’t much after training the model in the lecture so to understand completely what should I refer to do well in this course…??
Yes but you need to create the pipeline. The purpose of the project is understand the various operations and when to use those operations.
In training the dataset we just segregate one column for the train set and we train the set using two datasets like why are we segregating into two datasets, like on passing the segregated whole dataset are we training our model to predict single column which we are passing along with it…??
The column that try to predict is called the label. It is represented by “y”. And the training data is represented as X. we train the model using X and y. usually this is done by model.fit(X, y) method in sklearn. I hope I made sense.
To completely train a model first separate it into test and train set, next just pass it through the pipeline, next train the model and now if we fine-tune is that all that we have trained a model…??
Yes.
After training the model sir has not completely explained about fine-tuning the model as I am a beginner to the programming world i didn’t much after training the model in the lecture so to understand completely what should I refer to do well in this course…??
I would suggest doing the remaining projects. Also, try bootml.com.
1 Like
Thank you for taking your time and replying for me sir.
I have another doubt, in this we just import things from scikit learn and we run the program, I want to know where does the main work of a machine learning engineer lies sir if it is just importing things and running the program anyone can do know sir, there should be some task which is difficult where they require a machine learning engineer to complete it, what is that task sir till end to end you did not tell about it sir can you please explain sir.
Hi @Visaal_K.S,
It depends upon the profile of your job requirements. If someone has joined as a newbie in Machine Learning Division of some company, most of the projects are distributed among co-workers which mostly work on prototyping various parts of the project that has been assigned to them.
Most of the key decisions are taken by the Product Management or Head of the Project division of the company.
Currently, Machine Learning is in a steam engine phase in most industries including startups. So, its safe to assume that there is a lot of scope for improvement as the industry move towards production grade ML/AI systems being incorporated in their products itself apart from tech giants like Facebook, Google, Amazon and H2O which are way ahead in this field.
I hope it helps.
Regards