MESSAGE
| DATE | 2017-03-29 |
| FROM | Ruben Safir
|
| SUBJECT | Subject: [Learn] This is hard to understand what the logic is here
|
From learn-bounces-at-nylxs.com Wed Mar 29 01:31:31 2017
Return-Path:
X-Original-To: archive-at-mrbrklyn.com
Delivered-To: archive-at-mrbrklyn.com
Received: from www.mrbrklyn.com (www.mrbrklyn.com [96.57.23.82])
by mrbrklyn.com (Postfix) with ESMTP id 40C11161312;
Wed, 29 Mar 2017 01:31:31 -0400 (EDT)
X-Original-To: learn-at-nylxs.com
Delivered-To: learn-at-nylxs.com
Received: from [10.0.0.62] (flatbush.mrbrklyn.com [10.0.0.62])
by mrbrklyn.com (Postfix) with ESMTP id 3BD80160E77
for ; Wed, 29 Mar 2017 01:31:29 -0400 (EDT)
References: <430c32f2-da89-4077-b8a4-a5622fbb6b12.maildroid-at-localhost>
To: "learn-at-nylxs.com"
From: Ruben Safir
X-Forwarded-Message-Id: <430c32f2-da89-4077-b8a4-a5622fbb6b12.maildroid-at-localhost>
Message-ID: <712b3023-d054-3b79-5cc0-7bdb158e1e69-at-mrbrklyn.com>
Date: Wed, 29 Mar 2017 01:31:29 -0400
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:45.0) Gecko/20100101
Thunderbird/45.8.0
MIME-Version: 1.0
In-Reply-To: <430c32f2-da89-4077-b8a4-a5622fbb6b12.maildroid-at-localhost>
Subject: [Learn] This is hard to understand what the logic is here
X-BeenThere: learn-at-nylxs.com
X-Mailman-Version: 2.1.17
Precedence: list
List-Id:
List-Unsubscribe: ,
List-Archive:
List-Post:
List-Help:
List-Subscribe: ,
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Errors-To: learn-bounces-at-nylxs.com
Sender: "Learn"
There is a bias included here.
The bias is treated as a weight corresponding to a "pseudo" input that
is always 1. So for a perceptron with 2 regular inputs, the weight
vector will have size 3.
That while loop just appends ones to the input vector until it's big
enough to do a dot product of the inputs and weights.
The step function returns 0 or 1 but represents them as float because we
might want to substitute a different function, such as sigmoid, which
would need floats.
CL
-----Original Message-----
From: Ruben Safir
To: Christopher League , "learn-at-nylxs.com"
Sent: Wed, 29 Mar 2017 1:16
Subject: This is hard to understand what the logic is here
float feed_forward(const perceptron& p, vector& input)
{
// If input size doesn't match, add "bias" inputs:
while(input.size() < p.weights.size()) {
input.push_back(1);
}
return step_function(p.threshold, dot_product(p.weights, input));
}
step_function should return an INT
What is the logic to this? Don't you always need the bias?
https://gitlab.liu.edu/cs691s17/public/raw/1dd68b2f00b96db17a988242221f50e4eab7ddbb/perceptrons/perceptrons.cpp
_______________________________________________
Learn mailing list
Learn-at-nylxs.com
http://lists.mrbrklyn.com/mailman/listinfo/learn
|
|
