Topic Modeling Text Documents With LDA: Python in Excel Tutorial (Free Files)
Vložit
- čas přidán 23. 07. 2024
- ⬇️ Get the files and follow along: bit.ly/3QlcW3q
Topic modeling with Latent Dirichlet Allocation (LDA) allows you to extract information from your text documents!
It doesn't matter if you have emails, SMS text messages, Customer Service chats, or free-form fields in an IT system.
LDA topic modeling is useful to ANY professional.
☕ If you found this content useful and would like to support the channel, you can buy me a coffee: www.buymeacoffee.com/DaveOnData
--------------------------------------------------------------------------------------------
LDA ALGORITHM TUTORIALS
--------------------------------------------------------------------------------------------
Latent Dirichlet Allocation (Part 1 of 2):
• Latent Dirichlet Alloc...
Training Latent Dirichlet Allocation: Gibbs Sampling (Part 2 of 2):
• Training Latent Dirich...
--------------------------------------------------------------------------------------------
VIDEO CHAPTERS
--------------------------------------------------------------------------------------------
00:00 Intro
01:10 The BBC Dataset
02:08 Processing the Text Data
06:34 Training the LDA Topic Model
09:00 Which Docs Have Which Topics?
10:15 Which Words Belong to Which Topics?
11:52 How Many Topics?
14:45 What’s Next?
#pythoninexcel #pythonexcel #pythonforexcel - Věda a technologie
Incredibly explained video. Super professional but also accessible
Wow! Thank you so much for the kind words. Glad you enjoyed the content.
Text analytics has got my attention. Still waiting on my extraction at work and can't wait. Got the approval.
Woohoo! Text analytics is not only super useful, it's a lot of fun as well.
Hello! I don’t trust the neural network, and I’m trying to build the algorithm by myself and understand its structure, but it takes too much time, should I understand how the neural network works or should I just trust it looking at the result? What's the problem here?
This is a tradeoff common to many of the most useful machine learning algorithms-they are hard (or impossible) to interpret directly. Multilayer neural networks are a prime example of this.
Techniques like permutation importance and SHAP can help you understand the model's behavior/prediction logic, but they will never reach the level of interpretability that you get using something like logistic regression.
Is it possible to increase the smartness of a chat bot or neural network? Does it depend on what information is fed to him? For example, if I feed him a textbook on quantum mechanics, will he become smarter? Or does it depend on the initial AI parameters? How is this regulated?
In general, ML models are mostly the function of the data provided to train them.
In theory, you could train specialize models if you have enough source data.
To be honest, this is not my area of expertise as most of my clients benefit most from "traditional" ML.
@@DaveOnData Python, which is integrated into Excel, is capable of creating complex functions from several ones? Is the functionality the same as Python or different?