Use correct naming for iteration and epoch
This commit is contained in:
@@ -7,7 +7,6 @@ use Illuminate\Broadcasting\InteractsWithSockets;
|
||||
use Illuminate\Contracts\Broadcasting\ShouldBroadcast;
|
||||
use Illuminate\Foundation\Events\Dispatchable;
|
||||
use Illuminate\Queue\SerializesModels;
|
||||
use Illuminate\Support\Facades\Log;
|
||||
|
||||
class PerceptronTrainingIteration implements ShouldBroadcast
|
||||
{
|
||||
@@ -17,7 +16,7 @@ class PerceptronTrainingIteration implements ShouldBroadcast
|
||||
* Create a new event instance.
|
||||
*/
|
||||
public function __construct(
|
||||
public array $iterations, // ["iteration" => int, "exampleIndex" => int, "error" => float, "synaptic_weights" => array]
|
||||
public array $iterations, // ["epoch" => int, "exampleIndex" => int, "error" => float, "synaptic_weights" => array]
|
||||
public string $sessionId,
|
||||
public string $trainingId,
|
||||
)
|
||||
|
||||
@@ -4,6 +4,7 @@ namespace App\Models;
|
||||
|
||||
use App\Events\PerceptronTrainingEnded;
|
||||
use App\Services\DataSetReader;
|
||||
use App\Services\IPerceptronIterationEventBuffer;
|
||||
use App\Services\ISynapticWeightsProvider;
|
||||
use App\Services\PerceptronIterationEventBuffer;
|
||||
|
||||
@@ -18,36 +19,36 @@ class GradientDescentPerceptronTraining extends NetworkTraining
|
||||
public function __construct(
|
||||
DataSetReader $datasetReader,
|
||||
protected float $learningRate,
|
||||
int $maxIterations,
|
||||
int $maxEpochs,
|
||||
protected ISynapticWeightsProvider $synapticWeightsProvider,
|
||||
PerceptronIterationEventBuffer $iterationEventBuffer,
|
||||
IPerceptronIterationEventBuffer $iterationEventBuffer,
|
||||
string $sessionId,
|
||||
string $trainingId,
|
||||
private float $minError,
|
||||
) {
|
||||
parent::__construct($datasetReader, $maxIterations, $iterationEventBuffer, $sessionId, $trainingId);
|
||||
parent::__construct($datasetReader, $maxEpochs, $iterationEventBuffer, $sessionId, $trainingId);
|
||||
$this->perceptron = new GradientDescentPerceptron($synapticWeightsProvider->generate($datasetReader->getInputSize()));
|
||||
}
|
||||
|
||||
public function start(): void
|
||||
{
|
||||
$this->iteration = 0;
|
||||
$this->epoch = 0;
|
||||
do {
|
||||
$this->epochError = 0;
|
||||
$iterationErrorPerWeight = [];
|
||||
$this->iteration++;
|
||||
$epochCorrectorPerWeight = [];
|
||||
$this->epoch++;
|
||||
|
||||
while ($nextRow = $this->datasetReader->getRandomLine()) {
|
||||
$inputs = array_slice($nextRow, 0, -1);
|
||||
$correctOutput = (float) end($nextRow);
|
||||
|
||||
$iterationError = $this->iterationFunction($inputs, $correctOutput);
|
||||
$this->epochError += (1 / 2) * (abs($iterationError) ** 2); // TDDO REMOVEME abs()
|
||||
$this->epochError += ($iterationError ** 2) / 2;
|
||||
|
||||
// Store the iteration error for each weight
|
||||
$inputs_with_bias = array_merge([1], $inputs); // Add bias input
|
||||
foreach ($inputs_with_bias as $index => $input) {
|
||||
$iterationErrorPerWeight[$index][] = $iterationError * $input;
|
||||
$epochCorrectorPerWeight[$index][] = $iterationError * $input;
|
||||
}
|
||||
|
||||
// Broadcast the training iteration event
|
||||
@@ -57,14 +58,14 @@ class GradientDescentPerceptronTraining extends NetworkTraining
|
||||
// Synaptic weights correction after each epoch
|
||||
$synaptic_weights = $this->perceptron->getSynapticWeights();
|
||||
$new_weights = array_map(
|
||||
fn($weight, $weightIndex) => $weight + $this->learningRate * array_sum($iterationErrorPerWeight[$weightIndex]),
|
||||
fn($weight, $weightIndex) => $weight + $this->learningRate * array_sum($epochCorrectorPerWeight[$weightIndex]),
|
||||
$synaptic_weights,
|
||||
array_keys($synaptic_weights)
|
||||
);
|
||||
$this->perceptron->setSynapticWeights($new_weights);
|
||||
|
||||
$this->datasetReader->reset(); // Reset the dataset for the next iteration
|
||||
} while ($this->iteration < $this->maxIterations && !$this->stopCondition());
|
||||
} while ($this->epoch < $this->maxEpochs && !$this->stopCondition());
|
||||
|
||||
$this->iterationEventBuffer->flush(); // Ensure all iterations are sent to the frontend
|
||||
|
||||
|
||||
@@ -4,11 +4,11 @@ namespace App\Models;
|
||||
|
||||
use App\Events\PerceptronTrainingEnded;
|
||||
use App\Services\DataSetReader;
|
||||
use App\Services\PerceptronIterationEventBuffer;
|
||||
use App\Services\IPerceptronIterationEventBuffer;
|
||||
|
||||
abstract class NetworkTraining
|
||||
{
|
||||
protected int $iteration = 0;
|
||||
protected int $epoch = 0;
|
||||
|
||||
/**
|
||||
* @abstract
|
||||
@@ -18,8 +18,8 @@ abstract class NetworkTraining
|
||||
|
||||
public function __construct(
|
||||
protected DataSetReader $datasetReader,
|
||||
protected int $maxIterations,
|
||||
protected PerceptronIterationEventBuffer $iterationEventBuffer,
|
||||
protected int $maxEpochs,
|
||||
protected IPerceptronIterationEventBuffer $iterationEventBuffer,
|
||||
protected string $sessionId,
|
||||
protected string $trainingId,
|
||||
) {
|
||||
@@ -29,8 +29,8 @@ abstract class NetworkTraining
|
||||
abstract protected function stopCondition(): bool;
|
||||
|
||||
protected function checkPassedMaxIterations(?float $finalError) {
|
||||
if ($this->iteration >= $this->maxIterations) {
|
||||
$message = 'Le nombre maximal d\'itérations a été atteint';
|
||||
if ($this->epoch >= $this->maxEpochs) {
|
||||
$message = 'Le nombre maximal d\'epoch a été atteint';
|
||||
if ($finalError) {
|
||||
$message .= " avec une erreur finale de $finalError";
|
||||
}
|
||||
@@ -40,6 +40,6 @@ abstract class NetworkTraining
|
||||
}
|
||||
|
||||
protected function addIterationToBuffer(float $error, array $synapticWeights) {
|
||||
$this->iterationEventBuffer->addIteration($this->iteration, $this->datasetReader->getLastReadLineIndex(), $error, $synapticWeights);
|
||||
$this->iterationEventBuffer->addIteration($this->epoch, $this->datasetReader->getLastReadLineIndex(), $error, $synapticWeights);
|
||||
}
|
||||
}
|
||||
|
||||
@@ -4,8 +4,8 @@ namespace App\Models;
|
||||
|
||||
use App\Events\PerceptronTrainingEnded;
|
||||
use App\Services\DataSetReader;
|
||||
use App\Services\IPerceptronIterationEventBuffer;
|
||||
use App\Services\ISynapticWeightsProvider;
|
||||
use App\Services\PerceptronIterationEventBuffer;
|
||||
|
||||
class SimpleBinaryPerceptronTraining extends NetworkTraining
|
||||
{
|
||||
@@ -19,25 +19,25 @@ class SimpleBinaryPerceptronTraining extends NetworkTraining
|
||||
public function __construct(
|
||||
DataSetReader $datasetReader,
|
||||
protected float $learningRate,
|
||||
int $maxIterations,
|
||||
int $maxEpochs,
|
||||
protected ISynapticWeightsProvider $synapticWeightsProvider,
|
||||
PerceptronIterationEventBuffer $iterationEventBuffer,
|
||||
IPerceptronIterationEventBuffer $iterationEventBuffer,
|
||||
string $sessionId,
|
||||
string $trainingId,
|
||||
) {
|
||||
parent::__construct($datasetReader, $maxIterations, $iterationEventBuffer, $sessionId, $trainingId);
|
||||
parent::__construct($datasetReader, $maxEpochs, $iterationEventBuffer, $sessionId, $trainingId);
|
||||
$this->perceptron = new SimpleBinaryPerceptron($synapticWeightsProvider->generate($datasetReader->getInputSize()));
|
||||
}
|
||||
|
||||
public function start(): void
|
||||
{
|
||||
$this->iteration = 0;
|
||||
$this->epoch = 0;
|
||||
$error = 0;
|
||||
do {
|
||||
$this->iterationErrorCounter = 0;
|
||||
$this->iteration++;
|
||||
$this->epoch++;
|
||||
|
||||
while ($nextRow = $this->datasetReader->getRandomLine()) {
|
||||
while ($nextRow = $this->datasetReader->getNextLine()) {
|
||||
$inputs = array_slice($nextRow, 0, -1);
|
||||
$correctOutput = (float) end($nextRow);
|
||||
$correctOutput = $correctOutput > 0 ? 1 : 0; // Modify labels for non binary datasets
|
||||
@@ -48,7 +48,7 @@ class SimpleBinaryPerceptronTraining extends NetworkTraining
|
||||
$this->addIterationToBuffer($error, [[$this->perceptron->getSynapticWeights()]]);
|
||||
}
|
||||
$this->datasetReader->reset(); // Reset the dataset for the next iteration
|
||||
} while ($this->iteration < $this->maxIterations && !$this->stopCondition());
|
||||
} while ($this->epoch < $this->maxEpochs && !$this->stopCondition());
|
||||
|
||||
$this->iterationEventBuffer->flush(); // Ensure all iterations are sent to the frontend
|
||||
|
||||
|
||||
@@ -39,8 +39,8 @@ function getPerceptronErrorsPerIteration(): ChartData<
|
||||
dataset.data.push(iteration.error);
|
||||
|
||||
// Epoch error
|
||||
epochAverageError[iteration.iteration - 1] =
|
||||
(epochAverageError[iteration.iteration - 1] || 0) +
|
||||
epochAverageError[iteration.epoch - 1] =
|
||||
(epochAverageError[iteration.epoch - 1] || 0) +
|
||||
iteration.error ** 2 / 2;
|
||||
});
|
||||
|
||||
@@ -81,7 +81,7 @@ function getPerceptronErrorsPerIteration(): ChartData<
|
||||
plugins: {
|
||||
title: {
|
||||
display: true,
|
||||
text: 'Nombre d\'erreurs par itération',
|
||||
text: 'Nombre d\'erreurs par epoch',
|
||||
},
|
||||
},
|
||||
scales: {
|
||||
@@ -106,8 +106,8 @@ function getPerceptronErrorsPerIteration(): ChartData<
|
||||
}"
|
||||
:data="{
|
||||
labels: props.iterations.reduce((labels, iteration) => {
|
||||
if (!labels.includes(`Itération ${iteration.iteration}`)) {
|
||||
labels.push(`Itération ${iteration.iteration}`);
|
||||
if (!labels.includes(`Époch ${iteration.epoch}`)) {
|
||||
labels.push(`Époch ${iteration.epoch}`);
|
||||
}
|
||||
return labels;
|
||||
}, [] as string[]),
|
||||
|
||||
@@ -202,7 +202,7 @@ watch(selectedDatasetCopy, (newValue) => {
|
||||
<!-- MAX ITERATIONS -->
|
||||
<FormField name="max_iterations">
|
||||
<FormItem>
|
||||
<FormLabel>Nombre maximum d'epoch</FormLabel>
|
||||
<FormLabel>Nombre maximum d'itérations</FormLabel>
|
||||
<FormControl>
|
||||
<Input
|
||||
type="number"
|
||||
|
||||
@@ -1,7 +1,7 @@
|
||||
import type { Point } from "chart.js";
|
||||
|
||||
export type Iteration = {
|
||||
iteration: number;
|
||||
epoch: number;
|
||||
exampleIndex: number;
|
||||
weights: number[][][];
|
||||
error: number;
|
||||
@@ -10,7 +10,8 @@ export type Iteration = {
|
||||
export type Dataset = {
|
||||
label: string;
|
||||
data: DatasetPoint[];
|
||||
defaultLearningRate: number | undefined;
|
||||
defaultLearningRate?: number;
|
||||
defaultMinError?: number;
|
||||
};
|
||||
|
||||
export type DatasetPoint = {
|
||||
|
||||
Reference in New Issue
Block a user