Project 1 : SONAR Rock vs Mine Prediction with Python | End To End Python Machine Learning Project

[@Siddhardhan]

Hi! You can join this Telegram group for regular updates about my videos: t.me/siddhardhan Thank you!

[@uppariamericavlogsjustforf3395]

Hi Mr siddhardhan, I hope you doing well. I am unable to access your Google drive dataset .If you don’t mind please forward dataset of sonar rock vs mine prediction to [e-mail]. Thank you,

[@Siddhardhan]

Dataset sent

[@ggm4857]

already joined, nice channel.

[@paevents8064]

@@Siddhardhan @srii The data source is missing column heading sir

[@hemanthm1637]

sir in your previous video about types of supervised learning you said regression is used when it is about predicting quantity or continuous values like salary, age and price

[@RH-hv4ir]

First of all this project is so interesting!!! You are amazing, teaching means telling me why exactly we are we doing what we are doing. Thank you for being a real tutor and not being boring too! i was always questioning my brains ability but your way of teaching made me realize that its not my brain its the way we are being taught. I bet your channel is going to bring me to the next level

[@Siddhardhan]

Let's make a community of Machine Learning Experts! Subscribe here👉 tinyurl.com/md0gjbis

[@shiny123john]

I am planning to do mechine learning as a freelancer .Your video is very very useful to me. Thanks

[@camilleqdr]

Hi! I'm a french student and I just found your youtube channel, your videos are really great and useful, thank you so much !!

[@Siddhardhan]

Thanks a lot for your positive words 😇

[@lathaprasad1947]

Hey!! It would be great if you could share the documentation and code for this project. Please help your friend out.

[@brijeshnenwani]

@@lathaprasad1947 He has shared the google colab dataset and code in the description.

[@santoshhonnungar5543]

Camille.. i want to learn French.. please suggest me how to learn

[@01_abhijeet49]

@@santoshhonnungar5543 don't

[@extremexplorer8930]

This is Best ML Channel on KZbin....Peoples don't wanna see Real Thing............They just see Appeling Claikbaits on KZbin and Tries to Learn ML from those Guys..................

[@avatartwok588]

I've watched many tutorials on just learning general concepts, and this was y far the most comprehensive, easiest to understand video I've watched. Thank you, you and these videos are amazing.

[@aishanimaheshwari390]

your course is underrated and brilliant, god bless you sir!!!

[@dhruvilmalvania7035]

Good Going @Siddhardhan you really are doing this so good. I legit found this the best channel for my study purpose

[@MuhammadKamran-ii4rh]

So nice dear a great explanation....Now I am your fan from Indonesia. I am the student of master in mining engineering and I find this video so interesting. I will apply this to one of my model soon. I have not enough words to thank you enough

[@Siddhardhan]

Thank you so much 😇 I'll work constantly to add value to all the viewers of my channel 🤝

[@29ibrahimsayed95]

Assalamualaykum brother can you share your instagram or facebook i need to contact you ASAP

[@maheshrajbhar90]

Thanks, @Siddharthan bro! I had read about ML models before, but the concepts were pretty confusing. Now, I have a clear understanding of each operation-why we split into training and test sets, and how to measure the performance of the model. Thank you for the video! Keep up the great work!

[@rumeefeduma5596]

Thanks for this! Really great video, very intuitive even for beginner/intermediate python users.

[@Siddhardhan]

😇😇

[@yashu_goel]

Great explanation 👏👏👏👏 Liked and subscribed

[@smartlearnteam]

I just suscribed! Your content is really helpful to start doing projects which are so needed to develop skills on this field. Keep it up my friend.

[@c4yourselfyt]

Excellent video...great work boss !!!

[@Ndbdbbdbdbh]

This is some real good step by step explaination. Thanks for all your good work!

[@livetoyourfullest-n8b]

I think you might have interchanged the accuracy score parameters in this video, but otherwise, this is really amazing!!!thank you for helping me scale up my ML skills. Could you please do more videos on computer vision?

[@datascienceworld]

The video is highly informative. Just a small issue I am facing, the volume of the sound is low. However, thank you so much for your effort. Great help, indeed.

[@OpeyemiOlasunkanmi-mv1pz]

thanks sir for this work. you are blessed. you have really contributed something huge to your generation. we appreciate it.

[@benyusu8045]

Amazing

[@mdsohrabakhtaremam9555]

I am looking for this type of videos in which full and clear explanations on Machine learnings.. Thanks a lot

[@Siddhardhan]

You are welcome

[@yohannesayana9456]

Your presentation is priceless...just incredible! keep doing such wonderful tutorials.

[@Siddhardhan]

Thanks a ton!😇

[@hakantekin2701]

Your videos terrific and very helpful. Your explanation is very clear and understandable. Thank you so much

[@syedz9072]

Bro, your Work is fantastic and loving it. Very Much Appreciated

[@hamzaker829]

This is soooooo good; Blessings buddy.

[@danielsehnoutek2016]

great tutorial, love from india

[@shashankag5361]

superb bro, you are great instructor, i had never seen such explanatory video.

[@Siddhardhan]

Thanks a ton😇

[@KumR]

Wow... never thought about such a project. Super Thambi.

[@siyays1868]

Thanku so much Siddharthan for a wonderful machine learning project video. Your channel is really very good n videos r really great & one can get clearity about machine learning projects basics easily. Waiting for more such videos.....🥰

[@shanr8399]

Great explanation. Looking forward to learn as much as possible.

[@Siddhardhan]

Glad it was helpful!😇

[@asimrafidhaque9984]

Very good tutorial. Thanks!

[@dineshjothiram3677]

Organized Delivery.Excellent sir!

[@Siddhardhan]

Thanks a lot 😇

[@sokiynanaser5034]

Well explained. Great Job!👏👏👏

[@HarshitaChattopadhyay]

*We need more such videos related to projects in MACHINE LEARNING*

[@saketml8676]

Sir thankyou so much for your machine learning course ,your teaching style is fantastic , I was really confused regarding the data sets and proper model working ,you really cleared my all doubts ,thank you so much sir

[@Siddhardhan]

You are most welcome😇

[@nehakaler6677]

Just completed the implementation. Great video, thankyou.

[@divyanshupant3892]

Thankyou for sharing your project. Due to your video I just now know how to apply my theory knowledge into practical approach. I would have loved it there were graphs also for logistic Regression understanding for beginners. Thankyou so much.

[@lecturesfromleeds614]

You have a really good channel!

[@vaibhavparkale1113]

thank you so much bro for such a awesome video , keep it up .....

[@Siddhardhan]

Most welcome 😊

[@techshark7194]

Excellent!!want more intermediate to advanced use cases.

[@Siddhardhan]

Thank you so much! I'll be posting project videos every Friday. Stay tuned! Monday and Wednesday will be basic videos for beginners. Thanks for your support!

[@yashu_goel]

Please mention why are you using a particular algo.

[@yashwanthnamburi3824]

because it's binary classification and we use logistic regression when data is linearly separable and also interpretability is important

[@sach2274]

Amazing video. Loved it❤

[@alioraqsa]

This video is so easy to understand!!

[@hirk_IIT_BHU]

so well explained brother i was so confused with all these libraries & funcs...

[@NinadNakhwa]

At 22:11 there is function which finds mean of Y variable " sonar_data.groupby(60).mean() " , How can mean of existing dataset which already have resut will be usefull to predict on dataseet which does not contain Y predict.

[@Siddhardhan]

hi! it's not the result of the data. we are just exploring the data. we are just doing some data analysis. in this case, we are clearly seeing the difference in the mean value. but it's not the way the model understands the data. it tries to fit to the data and learn iteratively. we cannot create a model mentioning the mean value for all the columns. then it's explicitly telling the model about the data. and moreover, we cannot find this difference in mean in all datasets.

[@ishrakhamim]

You are an amazing instructor sir!

[@shivamSinha-iq1xz]

Very nice tutorial , u r helping me alot thank you so much 🎉

[@hizz4490]

Found very useful, thanks and keep up the work

[@dikshagupta3276]

Nice explanation big thanku to you pls make more and more videos related to data science project and provide some guidance...

[@hridhaanfilms2415]

Hello!! Siddhardhan I have subscribed to your channel and accessing the videos on Machine Learning. The videos are very informative and precise. I would like to know about some certifications for Machine Learning, so that I can get a job or use it for my higher studies

[@Khaled_Elsadani]

Thank you wish you all the best👌

[@sandipansarkar9211]

finished practicing coding.Feeling a lot more confident

[@antruong8770]

I wish I can catch what you are talking about, you have a very unique accent...

[@krshah2008]

What a wonderful video and the complete playlist as well. I was looking for something like this to improve my ML skills! One question - around 36:00, we see the model accuracy on training data to be about 83%... Shouldn't that be 100%? As we have created the model using the training data, and we are doing the prediction on the same data as well?

[@Siddhardhan]

thanks a lot 😇 we won't get 100% accuracy all the time. we may need to do some model optimization to make better predictions. you can research about it.

[@mohamedhamiche]

@@Siddhardhan Thank's for your work, I'm wondering why should we compute the accuracy score on training data, I mean for what purpose ? I don't think it is usefulll

[@piewpok3127]

@@mohamedhamiche yeah same.

[@rakeshdey6970]

very good.. Request you give more theoretical concept on each algorithm topic, it will be easier for us then to understand the usecase and practical

[@Siddhardhan]

Hi Rakesh! Thanks for your appreciation!😇 I'll definitely make detailed videos on the theory behind important Machine Learning models. But I cannot do it in these project videos. I'll make a module separately in my machine learning course in this KZbin channel, in which I'll explain about all the models in detail.

[@rakeshdey6970]

@@Siddhardhan Thanks.. Also want to know how we will approach if labelled data is not divided properly, In you example like R-111, M-97.. if those value not closed then how we will approach.. I am beginner so I realy like this demos. Thanks a lot

[@Siddhardhan]

We can use methods like under-sampling and over-sampling. In under-sampling, we reduce the labels that are more abundant and choose the important data points that are unique. In over-sampling, we try to make new data points by analysing the data with low number of labels. We can use algorithms like Bootstrapping or Synthetic Minority Over-sampling for this purpose.

[@Siddhardhan]

I hope this clears your doubt.

[@rakeshdey6970]

@@Siddhardhan Thanks , yes theoratically it is cleared, but want one short demo on over samplimg and under sampling when you will get time.. 😀

[@pranavip999]

Sir, Which is better Anaconda Jupyter notebook or google colaboratory

[@mike777med7]

Really cool content !!! 😁👍

[@Siddhardhan]

thanks a lot 😇

[@uchegodswill-iv4cd]

good lecture, i like this and i understood very well

[@satyapujari7731]

Loved the way explained.

[@shah.md.mostakim]

Such a great teacher 🔥

[@e-queen3485]

thank u sir it was soo helpfull and easy to understand

[@abhiramipr195]

Excellent ..thanku so much bro

[@safistya4254]

Well done sir

[@shashpeiris6008]

Your explanations are very clear. Can you please do some tutorials on probability pls? I have big issue in that

[@alfattahsuyadi3618]

Thank you very much. Very useful tutorial

[@its-itish]

Thank you so much ! wonderful lecture..

[@pahadiEngineer]

Real life working and advantage of this project....

[@yhbarve]

Wow! Nice video! And also learnt about Google Collab which was great. Thanks! Would love to connect with you

[@pawanputhran1816]

Could you please expand why was Logistics Regression used instead of any classification algorithms?

[@panagiotisgoulas8539]

The idea is to test many classification algorithms Knearest neighbors, Random Forests etc, calculate the acuracy on each of these, get an average accuracy for each of these after some iterations or with cross_val_score and finally you pick the one with with highest average accuracy.

[@prawinselvam9102]

Very useful video bro🔥

[@Siddhardhan]

Thanks bro😇

[@mudasiralisamejo8808]

Hey. One question about how should I approach these projects? Should I do exactly what is done in these videos or should I do things differently? I am just a beginner. Suggest me the best way for optimal learning and making the projects my own so that I can showcase them. Thank you for your time.

[@cherifihana9978]

i would recommend start by doing exactly what he is doing , one u feel yourself comfortable enough with the different concepts, u can start working or ur own

[@ganeshthapa3841]

what is the logic behind the data splitting into 1:9 or 2:8 train and test data? how to determine the test-size of the data?

[@ZeYe-u5q]

Thank you so much sir. You are so helpful!

[@abhishekpatil7336]

Good explanation bro🔥🔥

[@Siddhardhan]

Thanks brother!

[@LoneWolf-rj1px]

Great video to learn how to ML on Data Sets. I am struggling in the EDA and Data Preprocessing part!

[@debattambhattacharjee1642]

Really helpful for my projects.

[@Siddhardhan]

you're welcome

[@mihaluntinen2342]

Nice explanation of the coding and the function calls. I'm disappointed that there's no real discussion on what the data is (beyond "R = rock, M = mine", much less why a certain model is suitable to fit it. These are things that I was looking forward to learn about, and things that seem to me to be rather key elements of any ML project. But I'm sure this video is helpful to someone who already knows such things.

[@panther_.gaming]

here the target variable is binary class (2 values only).Such cases can be fitted by logistic regression model

[@dannymuzata4633]

Interesting video with crystal clear explanation. Thanks!! I have a question, in this video we found accuracy of test data to be ~76%. Is there anything that can be done (fine-tuning) to the dataset to improve this accuracy??? Also, is there a way we can show the failure side (24%) of prediction??? Are there any other models that can be used to solve a binary classification problem apart from LogReg??? HELP WITH THESE QUESTIONS. GREAT VIDEO!!!!!

[@GurpreetSingh-ps6kl]

i think we need more data for more accuracy

[@RamanaKumar-nn7fb]

Hi u have selected input_data why did u do that step and are the inputs taken random or particular

[@roybrodrick]

Take it this way, a company wants to check using some features to know if a Rock or Mine is under their submarine, so they enter 60 features to check it. If it returns R or M, then we know what it is. So that input_data section is just a random example from the dataset to see if it's correct

[@sohailsj8944]

That was a great tutorial

Thanks fot the exercise, I love it.

[@kummithasivamohanreddy6048]

Excellent 😍😍

[@Siddhardhan]

Thanks 🤗

[@satviksingh6229]

good concept, need some basic projects which beginners can add-in resume

[@Siddhardhan]

This is a very basic one... Try to learn this and execute the code by yourself. Will add more videos every Friday. Thanks!

[@tanzimhasan4464]

Thanks a lot for this Amazing video

[@Siddhardhan]

Most welcome 😊

[@stonecrane167]

do i need to know pytorch or tensorflow to do these projects

[@manthanmehar6792]

How you came up with the decision of using logistic regression?

[@DivyanshuSinghania01]

just a random decision we can use others too, in my case i used random forest and it had better result but might have some overfitting😅😅😭

[@deoshas520]

this is just great, i use it while learn ML i've got a problem actually: i don't see the results in 16th row - it's just "Logistic regression()" text with no results there's in video help

[@sukendar4316]

tq for the project can you explain 41:50 reshape part again? will you also upload data science projects?

[@Siddhardhan]

I'll make them in the future definitely! Now the reshape part: While training the model, we use the dataset which has 200 examples (rows) with 60 features (columns). Now in the prediction part, we are trying to predict only one instance. If we don't reshape the array, the model will think that we are again feeding 200 examples & it will give a error message due to this. That's why we reshape the input array. You get it?

[@ΓιώργοςΒαμβακούσης-ρ1τ]

Useful playlist, thank you!

[@Siddhardhan]

you're most welcome 😇

[@impulse1712]

Thank you so much sir..you explain very clearly..it helped a lot.😁

[@Siddhardhan]

Most welcome 😊

[@rafaelfraga7976]

I'm trying to do it on Jupyter Notebook and some things as sonar_data[60].value_counts and sonar_data.groupby(60),mean() isnt working

[@Pankaj_Khanal_Joshi]

hello sir in this example can you pls tell us which is dependent variable and independent variable. Thank you

[@demianaayman3802]

thanks sir for this work. you are blessed

[@uttamramanakatti4524]

Good explanation bro..

[@Siddhardhan]

Thank you so much!😇

[@baccouchekassem9575]

Sir can you please explain the 14:08 part how did you enclose the path in the quotes and put it in the brackets what button should I press to do it?

[@tejasbenibagde5725]

Use keyboard ⌨️😂, what else do you think everyone uses

[@anshtanwar2921]

your training data accuracy is more than test data, it should be a overfitting problem?

[@charanpoojary4804]

Thank you very much sir

[@VignithaAmpally-w8v]

Hi Siddhardhan. Thank you for the video. You mentioned about the difference in the Means making an impact in terms of prediction. How did we use that in this video?

[@binienjoyer]

Hello, which one do you think is the easiest to do for beginners with no background in these kind of topics? Our teacher is asking us to choose 4 projects from this video. Thank you!

[@ABBLACK-ls1cx]

WOW NICE BUT BRUH LITTLE SHORTER VIDEOS PLAYLIST PLEASE

[@todymaverick]

for the data sets why do you provide a drive link and not a kaggle one?

[@Ali-wo3rb]

If i would to this case by hand , how many inputs and hidden layers And the output shoul be 2 right ? M or R