forked from pms/uczenie-maszynowe
Wykłady 13-15 - poprawki
This commit is contained in:
Paweł Skórzewski
parent
13052aae72
commit
7f80072aea
@ -1,7 +1,6 @@
|
||||
{
|
||||
"cells": [
|
||||
{
|
||||
"attachments": {},
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"source": [
|
||||
@ -194,7 +193,7 @@
|
||||
"\n",
|
||||
"Żeby zredukować liczbę parametrów, a dzięki temu uprościć obliczenia, stosuje się warstwy ***pooling***.\n",
|
||||
"\n",
|
||||
"*Pooling* to rodzaj próbkowania. Najpopularniejszą jego odmianą jest *max-pooling*, czyli wybieranie najwyższej wartości spośród kilku sąsiadujących pikseli (rys. 12.1)."
|
||||
"*Pooling* to rodzaj próbkowania. Najpopularniejszą jego odmianą jest *max-pooling*, czyli wybieranie najwyższej wartości spośród kilku sąsiadujących pikseli (rys. 13.1)."
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -205,9 +204,9 @@
|
||||
}
|
||||
},
|
||||
"source": [
|
||||
"![Rys. 12.1. Pooling](Max_pooling.png \"Rys. 12.1. Pooling\")\n",
|
||||
"![Rys. 13.1. Pooling](Max_pooling.png \"Rys. 13.1. Pooling\")\n",
|
||||
"\n",
|
||||
"Rys. 12.1. - źródło: [Aphex34](https://commons.wikimedia.org/wiki/File:Max_pooling.png), [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0), Wikimedia Commons"
|
||||
"Rys. 13.1. - źródło: [Aphex34](https://commons.wikimedia.org/wiki/File:Max_pooling.png), [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0), Wikimedia Commons"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -218,7 +217,7 @@
|
||||
}
|
||||
},
|
||||
"source": [
|
||||
"Warstwy _pooling_ i konwolucyjne można przeplatać ze sobą (rys. 12.2)."
|
||||
"Warstwy _pooling_ i konwolucyjne można przeplatać ze sobą (rys. 13.2)."
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -229,9 +228,9 @@
|
||||
}
|
||||
},
|
||||
"source": [
|
||||
"![Rys. 12.2. CNN](Typical_cnn.png \"Rys. 12.2. CNN\")\n",
|
||||
"![Rys. 13.2. CNN](Typical_cnn.png \"Rys. 13.2. CNN\")\n",
|
||||
"\n",
|
||||
"Rys. 12.2. - źródło: [Aphex34](https://commons.wikimedia.org/wiki/File:Typical_cnn.png), [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0), Wikimedia Commons"
|
||||
"Rys. 13.2. - źródło: [Aphex34](https://commons.wikimedia.org/wiki/File:Typical_cnn.png), [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0), Wikimedia Commons"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -580,7 +579,7 @@
|
||||
"celltoolbar": "Slideshow",
|
||||
"email": "pawel.skorzewski@amu.edu.pl",
|
||||
"kernelspec": {
|
||||
"display_name": "Python 3",
|
||||
"display_name": "Python 3 (ipykernel)",
|
||||
"language": "python",
|
||||
"name": "python3"
|
||||
},
|
||||
@ -595,7 +594,7 @@
|
||||
"name": "python",
|
||||
"nbconvert_exporter": "python",
|
||||
"pygments_lexer": "ipython3",
|
||||
"version": "3.10.6 (main, Nov 14 2022, 16:10:14) [GCC 11.3.0]"
|
||||
"version": "3.10.6"
|
||||
},
|
||||
"livereveal": {
|
||||
"start_slideshow_at": "selected",
|
||||
|
||||
@ -1,17 +1,24 @@
|
||||
{
|
||||
"cells": [
|
||||
{
|
||||
"attachments": {},
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"metadata": {
|
||||
"slideshow": {
|
||||
"slide_type": "slide"
|
||||
}
|
||||
},
|
||||
"source": [
|
||||
"### Uczenie maszynowe\n",
|
||||
"# 14. Rekurencyjne sieci neuronowe"
|
||||
]
|
||||
},
|
||||
{
|
||||
"attachments": {},
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"metadata": {
|
||||
"slideshow": {
|
||||
"slide_type": "subslide"
|
||||
}
|
||||
},
|
||||
"source": [
|
||||
"## 14.1. Rekurencyjne sieci neuronowe"
|
||||
]
|
||||
@ -50,11 +57,11 @@
|
||||
"source": [
|
||||
"### Rekurencyjna sieć neuronowa – schemat\n",
|
||||
"\n",
|
||||
"Rys. 11.1.\n",
|
||||
"Rys. 14.1.\n",
|
||||
"\n",
|
||||
"![Rys. 11.1. Rekurencyjna sieć neuronowa – schemat](Recurrent_neural_network_unfold.png \"Rys. 11.1. Rekurencyjna sieć neuronowa – schemat\")\n",
|
||||
"![Rys. 14.1. Rekurencyjna sieć neuronowa – schemat](Recurrent_neural_network_unfold.png \"Rys. 14.1. Rekurencyjna sieć neuronowa – schemat\")\n",
|
||||
"\n",
|
||||
"Rys. 11.1 - źródło: [fdeloche](https://commons.wikimedia.org/wiki/File:Recurrent_neural_network_unfold.svg), [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0), Wikimedia Commons"
|
||||
"Rys. 14.1 - źródło: [fdeloche](https://commons.wikimedia.org/wiki/File:Recurrent_neural_network_unfold.svg), [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0), Wikimedia Commons"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -67,11 +74,11 @@
|
||||
"source": [
|
||||
"### LSTM – schemat\n",
|
||||
"\n",
|
||||
"Rys. 11.2.\n",
|
||||
"Rys. 14.2.\n",
|
||||
"\n",
|
||||
"![Rys. 11.2. LSTM – schemat](Long_Short-Term_Memory.png \"Rys. 11.2. LSTM – schemat\")\n",
|
||||
"![Rys. 14.2. LSTM – schemat](Long_Short-Term_Memory.png \"Rys. 14.2. LSTM – schemat\")\n",
|
||||
"\n",
|
||||
"Rys. 11.2 - źródło: [fdeloche](https://commons.wikimedia.org/wiki/File:Long_Short-Term_Memory.svg), [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0), Wikimedia Commons"
|
||||
"Rys. 14.2 - źródło: [fdeloche](https://commons.wikimedia.org/wiki/File:Long_Short-Term_Memory.svg), [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0), Wikimedia Commons"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -136,11 +143,11 @@
|
||||
"source": [
|
||||
"### GRU – schemat\n",
|
||||
"\n",
|
||||
"Rys. 11.3\n",
|
||||
"Rys. 14.3\n",
|
||||
"\n",
|
||||
"![Rys. 11.3. GRU – schemat](Gated_Recurrent_Unit.png \"Rys. 11.3. GRU – schemat\")\n",
|
||||
"![Rys. 14.3. GRU – schemat](Gated_Recurrent_Unit.png \"Rys. 14.3. GRU – schemat\")\n",
|
||||
"\n",
|
||||
"Rys. 11.3 - źródło: [fdeloche](https://commons.wikimedia.org/wiki/File:Gated_Recurrent_Unit.svg), [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0), Wikimedia Commons"
|
||||
"Rys. 14.3 - źródło: [fdeloche](https://commons.wikimedia.org/wiki/File:Gated_Recurrent_Unit.svg), [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0), Wikimedia Commons"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -158,7 +165,6 @@
|
||||
]
|
||||
},
|
||||
{
|
||||
"attachments": {},
|
||||
"cell_type": "markdown",
|
||||
"metadata": {
|
||||
"slideshow": {
|
||||
@ -197,7 +203,7 @@
|
||||
"* warstwa środkowa ma $k < n$ neuronów\n",
|
||||
"* $y^{(i)} = x^{(i)}$ dla każdego $i$\n",
|
||||
"\n",
|
||||
"(rys. 13.1)"
|
||||
"(rys. 14.4)"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -208,9 +214,9 @@
|
||||
}
|
||||
},
|
||||
"source": [
|
||||
"![Rys. 13.1. Autoencoder](Autoencoder_schema.png \"Rys. 13.1. Autoencoder\")\n",
|
||||
"![Rys. 14.4. Autoencoder](Autoencoder_schema.png \"Rys. 14.4. Autoencoder\")\n",
|
||||
"\n",
|
||||
"Rys. 13.1 - źródło: [Michela Massi](https://commons.wikimedia.org/wiki/File:Autoencoder_schema.png), [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0), Wikimedia Commons"
|
||||
"Rys. 14.4 - źródło: [Michela Massi](https://commons.wikimedia.org/wiki/File:Autoencoder_schema.png), [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0), Wikimedia Commons"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -240,7 +246,7 @@
|
||||
"* Ograniczenia nałożone na reprezentację danych w warstwie ukrytej pozwala na „odkrycie” pewnej **struktury** w danych.\n",
|
||||
"* _Decoder_ musi odtworzyć do pierwotnej postaci reprezentację danych skompresowaną przez _encoder_.\n",
|
||||
"\n",
|
||||
"(rys. 13.2)"
|
||||
"(rys. 14.5)"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -251,9 +257,9 @@
|
||||
}
|
||||
},
|
||||
"source": [
|
||||
"![Rys. 13.2. Autoencoder - struktura](Autoencoder_structure.png \"Rys. 13.2. Autoencoder - struktura\")\n",
|
||||
"![Rys. 14.5. Autoencoder - struktura](Autoencoder_structure.png \"Rys. 14.5. Autoencoder - struktura\")\n",
|
||||
"\n",
|
||||
"Rys. 13.2 - źródło: [Chervinskii](https://commons.wikimedia.org/wiki/File:Autoencoder_structure.png), [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0), Wikimedia Commons"
|
||||
"Rys. 14.5 - źródło: [Chervinskii](https://commons.wikimedia.org/wiki/File:Autoencoder_structure.png), [CC BY-SA 4.0](https://creativecommons.org/licenses/by-sa/4.0), Wikimedia Commons"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -302,7 +308,6 @@
|
||||
]
|
||||
},
|
||||
{
|
||||
"attachments": {},
|
||||
"cell_type": "markdown",
|
||||
"metadata": {
|
||||
"slideshow": {
|
||||
@ -438,7 +443,6 @@
|
||||
]
|
||||
},
|
||||
{
|
||||
"attachments": {},
|
||||
"cell_type": "markdown",
|
||||
"metadata": {
|
||||
"slideshow": {
|
||||
@ -481,7 +485,7 @@
|
||||
"celltoolbar": "Slideshow",
|
||||
"email": "pawel.skorzewski@amu.edu.pl",
|
||||
"kernelspec": {
|
||||
"display_name": "Python 3",
|
||||
"display_name": "Python 3 (ipykernel)",
|
||||
"language": "python",
|
||||
"name": "python3"
|
||||
},
|
||||
@ -496,7 +500,7 @@
|
||||
"name": "python",
|
||||
"nbconvert_exporter": "python",
|
||||
"pygments_lexer": "ipython3",
|
||||
"version": "3.10.6 (main, Nov 14 2022, 16:10:14) [GCC 11.3.0]"
|
||||
"version": "3.10.6"
|
||||
},
|
||||
"livereveal": {
|
||||
"start_slideshow_at": "selected",
|
||||
|
||||
@ -2,8 +2,13 @@
|
||||
"cells": [
|
||||
{
|
||||
"cell_type": "markdown",
|
||||
"metadata": {},
|
||||
"metadata": {
|
||||
"slideshow": {
|
||||
"slide_type": "slide"
|
||||
}
|
||||
},
|
||||
"source": [
|
||||
"### Uczenie maszynowe\n",
|
||||
"# 15. Uczenie przez wzmacnianie i systemy dialogowe"
|
||||
]
|
||||
},
|
||||
@ -15,7 +20,7 @@
|
||||
}
|
||||
},
|
||||
"source": [
|
||||
"# 15.1. Uczenie przez wzmacnianie"
|
||||
"## 15.1. Uczenie przez wzmacnianie"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -71,7 +76,7 @@
|
||||
}
|
||||
},
|
||||
"source": [
|
||||
"![Rys. 13.2. Agent i środowisko](agent_i_srodowisko.png \"Rys. 13.2. Agent i środowisko\")"
|
||||
"![Rys. 15.2. Agent i środowisko](agent_i_srodowisko.png \"Rys. 15.2. Agent i środowisko\")"
|
||||
]
|
||||
},
|
||||
{
|
||||
@ -257,9 +262,9 @@
|
||||
"source": [
|
||||
"## Architektura systemu dialogowego\n",
|
||||
"\n",
|
||||
"(rys. 13.3)\n",
|
||||
"(rys. 15.3)\n",
|
||||
"\n",
|
||||
"![Rys. 13.3. Architektura systemu dialogowego](system_dialogowy.png \"Rys. 13.3. Architektura systemu dialogowego\")"
|
||||
"![Rys. 15.3. Architektura systemu dialogowego](system_dialogowy.png \"Rys. 15.3. Architektura systemu dialogowego\")"
|
||||
]
|
||||
}
|
||||
],
|
||||
@ -268,7 +273,7 @@
|
||||
"celltoolbar": "Slideshow",
|
||||
"email": "pawel.skorzewski@amu.edu.pl",
|
||||
"kernelspec": {
|
||||
"display_name": "Python 3",
|
||||
"display_name": "Python 3 (ipykernel)",
|
||||
"language": "python",
|
||||
"name": "python3"
|
||||
},
|
||||
@ -283,7 +288,7 @@
|
||||
"name": "python",
|
||||
"nbconvert_exporter": "python",
|
||||
"pygments_lexer": "ipython3",
|
||||
"version": "3.10.6 (main, Nov 14 2022, 16:10:14) [GCC 11.3.0]"
|
||||
"version": "3.10.6"
|
||||
},
|
||||
"livereveal": {
|
||||
"start_slideshow_at": "selected",
|
||||
|
||||
BIN
wyk/Autoencoder_schema.png
Normal file
BIN
wyk/Autoencoder_schema.png
Normal file
Binary file not shown.
|
After Width: | Height: | Size: 39 KiB |
BIN
wyk/Autoencoder_structure.png
Normal file
BIN
wyk/Autoencoder_structure.png
Normal file
Binary file not shown.
|
After Width: | Height: | Size: 48 KiB |
BIN
wyk/Gated_Recurrent_Unit.png
Normal file
BIN
wyk/Gated_Recurrent_Unit.png
Normal file
Binary file not shown.
|
After Width: | Height: | Size: 53 KiB |
BIN
wyk/Long_Short-Term_Memory.png
Normal file
BIN
wyk/Long_Short-Term_Memory.png
Normal file
Binary file not shown.
|
After Width: | Height: | Size: 56 KiB |
BIN
wyk/Max_pooling.png
Normal file
BIN
wyk/Max_pooling.png
Normal file
Binary file not shown.
|
After Width: | Height: | Size: 25 KiB |
BIN
wyk/Recurrent_neural_network_unfold.png
Normal file
BIN
wyk/Recurrent_neural_network_unfold.png
Normal file
Binary file not shown.
|
After Width: | Height: | Size: 31 KiB |
BIN
wyk/Typical_cnn.png
Normal file
BIN
wyk/Typical_cnn.png
Normal file
Binary file not shown.
|
After Width: | Height: | Size: 116 KiB |
BIN
wyk/agent_i_srodowisko.png
Normal file
BIN
wyk/agent_i_srodowisko.png
Normal file
Binary file not shown.
|
After Width: | Height: | Size: 39 KiB |
BIN
wyk/paradygmaty_um.png
Normal file
BIN
wyk/paradygmaty_um.png
Normal file
Binary file not shown.
|
After Width: | Height: | Size: 63 KiB |
BIN
wyk/system_dialogowy.png
Normal file
BIN
wyk/system_dialogowy.png
Normal file
Binary file not shown.
|
After Width: | Height: | Size: 54 KiB |
Loading…
Reference in New Issue
Block a user