Hfrxdjt

Why must Chinese maps be obfuscated?

Is it acceptable to use working hours to read general interest books?

Is there any pythonic way to find average of specific tuple elements in array?

How bug prioritization works in agile projects vs non agile

Do I need to watch Ant-Man and the Wasp and Captain Marvel before watching Avengers: Endgame?

Are there moral objections to a life motivated purely by money? How to sway a person from this lifestyle?

Unknown code in script

What is the most expensive material in the world that could be used to create Pun-Pun's lute?

What makes accurate emulation of old systems a difficult task?

Negative Resistance

How important is it that $TERM is correct?

Partitioning values in a sequence

Multiple options vs single option UI

Is there really no use for MD5 anymore?

Can a Bard use the Spell Glyph option of the Glyph of Warding spell and cast a known spell into the glyph?

How do I check if a string is entirely made of the same substring?

Cayley's Matrix Notation

What is the term for a person whose job is to place products on shelves in stores?

Apply a different color ramp to subset of categorized symbols in QGIS?

All ASCII characters with a given bit count

"The cow" OR "a cow" OR "cows" in this context

Could moose/elk survive in the Amazon forest?

Co-worker works way more than he should

Can a stored procedure reference the database in which it is stored?

LDA for each target in a binary classification problem

0

$begingroup$

I have a question around using LDA which may be very foolish but I am still gonna ask anyway.

Problem Statement: To classify documents into complaints and queries by processing text.

Current Approach: After preprocessing text, I am using LDA on the entire text with 20 topics and feed the probabilities to a dataframe and use a classification algorithm to predict the classes.

Which I think is the correct way to use LDA

Second Approach: (This is where the confusion lies): Does it make sense to train complaints and queries separately with 10 topics using LDA and then construct a dataframe with 20 columns(10 from complaints, 10 from queries) to pass to a classification algorithm.

In the future, this model will have process an incoming text thru both the LDA models and pass the output to a classification algorithm in order to separate complaints and queries.

I don't believe that a second approach is a valid approach but someone has suggested me so and I am trying to understand how this could be valid. Or is there a way to take 10 topics from each complaint and query from the LDA model and feed them into a dataframe in order to predict if it is a complaint or a query.

classification nlp topic-model lda binary

share

asked 1 min ago

ShoaibkhanzShoaibkhanz

1012

$endgroup$

add a comment | 

0

$begingroup$

I have a question around using LDA which may be very foolish but I am still gonna ask anyway.

Problem Statement: To classify documents into complaints and queries by processing text.

Current Approach: After preprocessing text, I am using LDA on the entire text with 20 topics and feed the probabilities to a dataframe and use a classification algorithm to predict the classes.

Which I think is the correct way to use LDA

Second Approach: (This is where the confusion lies): Does it make sense to train complaints and queries separately with 10 topics using LDA and then construct a dataframe with 20 columns(10 from complaints, 10 from queries) to pass to a classification algorithm.

In the future, this model will have process an incoming text thru both the LDA models and pass the output to a classification algorithm in order to separate complaints and queries.

I don't believe that a second approach is a valid approach but someone has suggested me so and I am trying to understand how this could be valid. Or is there a way to take 10 topics from each complaint and query from the LDA model and feed them into a dataframe in order to predict if it is a complaint or a query.

classification nlp topic-model lda binary

share

asked 1 min ago

ShoaibkhanzShoaibkhanz

1012

$endgroup$

add a comment | 

$begingroup$

I have a question around using LDA which may be very foolish but I am still gonna ask anyway.

Problem Statement: To classify documents into complaints and queries by processing text.

Current Approach: After preprocessing text, I am using LDA on the entire text with 20 topics and feed the probabilities to a dataframe and use a classification algorithm to predict the classes.

Which I think is the correct way to use LDA

Second Approach: (This is where the confusion lies): Does it make sense to train complaints and queries separately with 10 topics using LDA and then construct a dataframe with 20 columns(10 from complaints, 10 from queries) to pass to a classification algorithm.

In the future, this model will have process an incoming text thru both the LDA models and pass the output to a classification algorithm in order to separate complaints and queries.

I don't believe that a second approach is a valid approach but someone has suggested me so and I am trying to understand how this could be valid. Or is there a way to take 10 topics from each complaint and query from the LDA model and feed them into a dataframe in order to predict if it is a complaint or a query.

classification nlp topic-model lda binary

share

asked 1 min ago

ShoaibkhanzShoaibkhanz

1012

$endgroup$

share

asked 1 min ago

ShoaibkhanzShoaibkhanz

1012

share

asked 1 min ago

ShoaibkhanzShoaibkhanz

1012

asked 1 min ago

ShoaibkhanzShoaibkhanz

1012

asked 1 min ago

ShoaibkhanzShoaibkhanz

1012