Training of two 3×3 convolution layers vs training one 5×5 convolution layer

I’m not 100% sure this is the right stackexchange, please feel free to redirect me to another one.

I know that two 3×3 convolution layers can be equivalent to one 5×5 convolution layer.

I also know that in many cases (maybe all of them), the training of both options is equivalent (in terms of result, not speed or optimisation)

Suppose I have a dataset on which I KNOW that no information whatsoever can be derived from any 3×3 window, but it can be from a 5×5 window.

If I wanted to, I could train a network with a 5×5 convolution layer, then manually transform it into two 3×3 convolution layers.

My question is: could I train a network directly with two 3×3 convolution layers ?

My reasoning: What could the first convolution learn apart from overfitting the training dataset? Supposedly nothing. So, the only layer that could some generalization of my data is the second one. But then, my problem sums up to comparing ONE 3×3 convolution versus ONE 5×5 convolution.

In the general case, the 5×5 convolution would perform better. But with my restrictions on the dataset, maybe it becomes equivalent to the 3×3 convolution? I which case I could train 3×3 convolution.

Thank you !

Training Documentation / Company Processes [on hold]

I work in a Saas company with a huge system, and set of processes. We have a wiki, lots of resources on google drive and information / documentation stored in a large number of random places (including some individuals’ own minds).

I am trying to develop a one stop shop knowledge base for our engineers…Initially the ideas was a resource that can be used for training new hires but as we started to discuss it, we realised how badly we need it to be the reference point for all existing employees. I work on a specific engineering team so will create it for our team initially and build out from there.

This is going to be a HUGE task as there is so much to gather together but I am looking for some recommendations to get started, specifically if anybody has good ideas / resources for how we could create this one stop shop and

  • Add system architecture diagrams
  • Include drill-down links to more detail and eventually down to code
  • List / link to tools available to engineers
  • A live area where individuals can add ‘troubleshooting tips’ (for issues they solved in our live environment)
  • Information on how our section of the system integrates with other teams

Or just ideas / recommendations on how these things are managed in companies you’ve worked for.

Simple Question About the Nature of Supervised Learning Predictive Models in Relation to their Training Data

Would a deployed or deployment-ready (i.e., already-trained) supervised learning model have a 100% accuracy rate for predictions on the training data (if you were to run the training data through it again)?

[Note: I don’t know if it is possible for such models to incorporate “memory” such that the model would retrieve an output variable for a given training data point (containing various input variables) from memory (having seen it before during training), instead of retrieving it by plugging the data point’s input variables into the equation that is the predictive model (obviously a unique, unknown equation created during the training process, with millions of parameters, which no human knows the precise nature of). If it is possible for such models to have “memory”, then for the purposes of my question, please assume it is not.]

What I would like to know is: are SL predictive models perfectly parameterized for the training data, such that feeding any training data point in to the completed model will 100% of the time yield the correct output for that data point (note: I am aware that, in general, testing a deployment-ready model on its training data is not a very useful test), or are such models imperfect predictors even if run (when already-trained) on their training data?

I know people talk about overfitting and the bias-variance tradeoff. That would seem to indicate the latter to be the case, but I want to be sure.

Thank you for any answers you can offer!

label your data for AI or ML training for $10

AI image classification models require training so that it can learn from those images, one of the most time-consuming jobs is to label the images in order to train the model. I will help you do that, I am offering a number of packages depending on the number of images you need to label. I will label 100 images for 10$ . Do contact me if you want to label more than 200 images.

by: hammad1997
Created: —
Category: Data Entry
Viewed: 276


500 Female Fitness Training Videos for $2

Hi, everybody. Thanks for choosing my Service. – MP4 format – 30 sec – 1 min – over 9 Gb – videos with owner links in the description. All videos on Google Drive. After order, i give you link to download files. Have you got any questions? I’m glad to help you. This videos will suit for you Instagram acc, facebook, blog, web site and etc. Ready to order? I am ready to work with you, so click the order button right now! With love! Best regards!

by: ciriaco1
Created: —
Category: Video
Viewed: 201


The Forex Training Guide. Expand your investment strategy with forex trading. for $10

Expand your investment strategy with forex trading by learning from the pros how to maximize your return on investment. There is sometimes over $ 4 trillion in average daily trading on the international foreign exchange market and you can get in on the action.Unlike the Nasdaq, Dow and other trading markets, the FOREX market is open 24 hours a day with a consistent flow of steady trading from around the world. It is a great opportunity for those who like to invest and earn in a market that never sleeps. But as a volatile market with a lot of options for investment gain or loss, learning the ropes is crucial. Get in on Some of the Action in the International Foreign Exchange Market and Start Making Profit Today with “The Forex Training Guide” This eBook will teach you everything you need to know to get started today, including: What the stock market is all about, stock market trends, and how to use this information in FOREX trading How you can make sure you have a firm grasp on the FOREX market through an insightful introduction A deeper understanding of currency trading, including its benefits and pitfalls How to deal with FOREX volatility and market expectation An in-depth look into trading statistics and how they apply to the FOREX market The various aspects of the FOREX trade and whether or not to invest long or short The way to deal with risk management, including working with limit orders and other risk minimization factors An explanation of all the FOREX buzz words so you can understand related discussion and investment terms And much, much more! Learn Effective FOREX Trading Strategy So You Don’t Lose Your Shirt! Don’t jump into international foreign exchange trading until you understand the dynamics. Many have lost their savings by trying to trade in the FOREX marketplace without a sound knowledge of what they are doing. Instead of learning the hard way, rely on the pros to show you how it is done. In other words, get valuable knowledge that will help you do it right and make money. Even if you have traded equities in the Nasdaq, the Dow or other markets, understanding the volatility and differences in the FOREX market are crucial if you want to make money instead of lose it. Let us explain it all to you and help you succeed.

by: abigadon
Created: —
Category: eBook
Viewed: 146


Issue with training text classification model in R on Ubuntu server 18.04

The R code I have is as following, it results an error as below. However, this works fine on my windows local machine. Just error on ubuntu server as below.The generated model.bin file is about 229 MB on my windows local machine. Not sure why I cannot train the model on Ubuntu server, any help would be appreciated!

library(fastrtext)  model_file <- build_supervised(documents = data1[["V5"]],                                targets = data1[["V3"]],                                model_path = '/srv/shiny-server/text_cls/voc_single/mymodel',                                dim = 30, lr = 0.1, epoch = 30, wordNgrams = 2)    Prepare and write tempfile at ... /tmp/RtmpLMf0uE/file9083d9d7650 Starting supervised training with following commands:  $   supervised -input /tmp/RtmpLMf0uE/file9083d9d7650 -output /srv/shiny-server/text_cls/voc_single/mymodel -lr 0.1 -dim 30 -ws 5 -epoch 30 -minCount 5 -minCountLabel 0 -neg 5 -wordNgrams 2 -loss ns -bucket 2000000 -minn 3 -maxn 6 -thread 12 -lrUpdateRate 100 -t 0.0001 -label __label__ -verbose 2   Error in model$  execute(c("fasttext", commands)) :    /srv/shiny-server/text_cls/voc_single/mymodel.bin cannot be opened for saving. 

[Get] Grant Cardone – Cardone University – Sales Training Rated 1 Of The Best

See Post # 23 For New DL Link:cool:

Hey Guys This Is Huge Share, No Whiners Or Complainers Allowed – Help Yourself.

Code:
 cardoneuniversity.com 

Download 100 Ways to Stay Motivated:
Closers Survival Guide
Closing Strategies
The 10x Rule
If you're not first, you're last
Incoming Calls
Inspiration from Grant
Prospecting Strategies
Quick Fix
Rules of Success
Sell or be Sold
Selling Basics
The…

[Get] Grant Cardone – Cardone University – Sales Training Rated 1 Of The Best