Emma, this is absolutely amazing. I just got a data engineering job where I’ll be working alongside data scientists on NLP problems, which I have zero background in. This video really shed light on the technical fundamentals of NLP. Thank you so much! Looking forward to your next videos.
Deborshi(pronounced as Deborkhi in our language)and I were born in the same state, just recognized my English accent similar to him, amazing, from our place, very people go on the advanced tech fields. I am trying to get good at AI as well, especially in deep learning.. I think I really love NLP, computer vision and tasks related to the RNN, LSTM, BERT and CNN architectures. As I am not a strong data scientist, just an AI developer, I think this channel is helping me to get familiarize with data science interviews, problems, and helped me to figure it out what things an employer actually expects from a good data scientist or AI developer. Thanks
Hi Emma, thanks for sharing with ua a lot of product sense questions and interview tips. I see that you are currently a software engineer, and I'm curious, with expertise in data science, what motivated you to switch to a SDE? can you share your thoughts in a video maybe :)
Thanks for sharing! After the whole video, I want to know, does anyone do the NLP project about embedding your resume into vector and comparing the other job description. And then find the closest one, maybe work somehow.
Your channel is a gem! One question though... I'm curious how comfortable you guys are with the mathematics behind these deep learning/machine learning algorithms. For example, support vector machines require quadratic programming and lagrangians -- are you asked about these kinds of stuff in interviews? Other algorithms get even more nasty as well o_O
Personally, while I think it's important to internalize basic ML theory, A. I've never been asked questions with that kind of depth during interviews (or at least in the roles that I've applied to in the past) B. I've never once had to think about quadratic programming and Lagrangians in my day to day job C. Core concepts in ML are things you should learn regardless of whether or not you're looking for NLP specific roles. I think the real value in knowing fundamentals is in being able to deliver high quality work that you are confident in as a data scientist. I think strong fundamentals enable that kind of rigor. But statistical/mathematical rigor is a part of the job and not all of it. They could ask you some of this stuff to test your fundamentals in ML. I think it is reasonable to ask a candidate to frame the optimization problem of SVM. Similarly, it is reasonable to ask someone to write the psuedo-code for kmeans. I think testing fundamentals is important, but I would never outright disqualify a candidate for not knowing the dual formulation of SVM (unless the job literally just requires you to optimize an SVM model and that's it).
00:02 NLP (Natural Language Processing) data science career advice 02:17 Journey from mechanical engineering to NLP 06:49 Transferable job skills 08:58 NLP & ML is challenging but offers opportunities for those with basic experience. 13:31 Text-to-number conversion methods in NLP 15:45 Foundational topics for data science 19:32 Free resources for learning data science and NLP. 21:30 Open source NLP data sets are available Share data science stories & and interview tips
