shaunrd0/klips
1
0
Fork 0
Code Issues Pull Requests 1 Projects Releases Wiki Activity

Add JavaScript WebGL example

Browse Source 
+ Update READMEs
This commit is contained in:
Shaun Reed
2022-02-07 13:55:14 -05:00
parent 940d035638
commit 573fc4e1e8
7 changed files with 191 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files

14
python/linear-regression/README.md
View File

@@ -32,20 +32,6 @@ optional arguments:
Optionally provide file for data to be read from. Each point must be on it's own line with format x,y
```
Running linear regression program
```bash
python3.9 linear-regression.py --file ./input.txt --silent
Finding fitting line plot for given data [(1, 3), (2, 7), (3, 5), (4, 9), (5, 11), (6, 12), (7, 15)]
points_avg: (5.117647058823529, 5.235294117647059)
variance: (241.76470588235296, 193.05882352941177)
sigma: (3.887196176892422, 3.4736402333270258)
covariance: 0.8455882352941174
correlation: 0.0626235432924427
Our line Y = BX + A must pass through the point (5.117647058823529, 5.235294117647059)
Y = (0.05596107055961069)X + 4.9489051094890515
For X = 4.5, Y is predicted to be 5.200729927007299
```
By default, the following linear regression is calculated and displayed
```bash
python3.9 linear-regression.py

2
python/markov-model/README.md
View File

@@ -90,7 +90,7 @@ Final paths sorted by probability:
(0, 2, 1, 0, 2) has probability: 0.001395
```
By default, a random Hidden Markov Model and visualization will be generated
By default, a random Hidden Markov Model and visualization will be generated and the sequence `[A, B, D, C]` will be used.
```bash
python3.9 markov-model.py

10
python/neural-network/README.md
View File

@@ -8,10 +8,10 @@ python3.9 -m pip install -r requirements.txt
Neural network implementation using Python CLI to dynamically generate a resizable network
and then run a given number of learning cycles on the provided data set.
As an example, the IRIS dataset is used to classify flower types using petal measurements.
Input layer perceptron count can be adjusted with `INPUTS` positional parameter
Hidden layer perceptron count can be adjusted with `PERCEPTRONS` positional parameter
Output layer perceptron count can be adjusted with `OUTPUTS` positional parameter
Hidden layers can be added or removed using`--hidden-layers` option setting
Input layer perceptron count can be adjusted with `INPUTS` positional parameter.
Hidden layer perceptron count can be adjusted with `PERCEPTRONS` positional parameter.
Output layer perceptron count can be adjusted with `OUTPUTS` positional parameter.
Hidden layers can be added or removed using`--hidden-layers` option setting.
Node bias can be initialized randomly or with provided data.
Perceptron edge weight bias can be initialized randomly or with provided data.
Threshold for perceptron fire can be initialized randomly or with provided data.
@@ -96,7 +96,7 @@ If the length of an input sequence does not match the number of input nodes requ
If the length of possible label classifications does not match the number of output nodes requested, a warning will show.
In both cases, the program corrects the node count to match the input data / labels, and not the requested node count.
The total number of output labels provided must match the total number of the number of input sequences.
The total number of output labels provided must match the total number of input sequences.
Running NN program uses IRIS data set by default.
Warnings will be shown if input and output node count is changed without providing new input.