https://wiki.mindmaker.it/index.php?action=history&feed=atom&title=The_Unreasonable_Effectiveness_of_Recurrent_Neural_Networks 2026-05-14T11:37:18Z Cronologia della pagina su questo sito MediaWiki 1.41.0 https://wiki.mindmaker.it/index.php?title=The_Unreasonable_Effectiveness_of_Recurrent_Neural_Networks&diff=2546&oldid=prev 2024-08-17T10:53:24Z

<p></p> <table style="background-color: #fff; color: #202122;" data-mw="interface"> <col class="diff-marker" /> <col class="diff-content" /> <col class="diff-marker" /> <col class="diff-content" /> <tr class="diff-title" lang="it"> <td colspan="2" style="background-color: #fff; color: #202122; text-align: center;">← Versione meno recente</td> <td colspan="2" style="background-color: #fff; color: #202122; text-align: center;">Versione delle 10:53, 17 ago 2024</td> </tr><tr><td colspan="2" class="diff-lineno" id="mw-diff-left-l1">Riga 1:</td> <td colspan="2" class="diff-lineno">Riga 1:</td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div><del style="font-weight: bold; text-decoration: none;">Titolo: [[Titolo::The Unreasonable Effectiveness of Recurrent Neural Networks]]</del></div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">{{template pubblicazione</ins></div></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div> </div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">|data=</ins>2015</div></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div><del style="font-weight: bold; text-decoration: none;">Anno di pubblicazione: [[AnnoDiPubblicazione::</del>2015<del style="font-weight: bold; text-decoration: none;">]]</del></div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">|autori=M. Fabbri, G. Moro</ins></div></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div> </div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">|</ins>URL<ins style="font-weight: bold; text-decoration: none;">=</ins>https://<ins style="font-weight: bold; text-decoration: none;">www</ins>.<ins style="font-weight: bold; text-decoration: none;">semanticscholar</ins>.<ins style="font-weight: bold; text-decoration: none;">org</ins>/<ins style="font-weight: bold; text-decoration: none;">paper</ins>/<ins style="font-weight: bold; text-decoration: none;">36c5bab50e0b8764bfdeefee09f9b1798686ed24</ins></div></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div><del style="font-weight: bold; text-decoration: none;">Autore: [[Autore::Andrej Karpathy]]</del></div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">|topic=RNN</ins></div></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div> </div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">|citazioni=25</ins></div></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div>URL<del style="font-weight: bold; text-decoration: none;">: [[URL::</del>https://<del style="font-weight: bold; text-decoration: none;">karpathy</del>.<del style="font-weight: bold; text-decoration: none;">github</del>.<del style="font-weight: bold; text-decoration: none;">io</del>/<del style="font-weight: bold; text-decoration: none;">2015</del>/<del style="font-weight: bold; text-decoration: none;">05/21/rnn-effectiveness/]]</del></div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">}}</ins></div></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div> </div></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">Sebbene le reti neurali ricorrenti (RNN) superino gli algoritmi di apprendimento automatico tradizionali nell'individuazione di dipendenze a lungo termine tra le istanze di addestramento, come nelle sequenze di termini nelle frasi o tra valori nelle serie temporali, sorprendentemente pochi studi finora hanno implementato soluzioni concrete con RNN per la negoziazione del mercato azionario. Presumibilmente le attuali difficoltà nell'addestramento </ins>delle RNN <ins style="font-weight: bold; text-decoration: none;">hanno contribuito a scoraggiarne l'ampia adozione. Questo lavoro presenta una soluzione semplice ma efficace, basata su una RNN profonda, i cui guadagni nel trading con Dow Jones Industrial Average (DJIA) superano lo stato dell'arte, inoltre il guadagno è superiore del 50% rispetto a quello prodotto da simili reti neurali profonde feed forward. Le azioni di trading sono guidate dalle previsioni dei movimenti di prezzo del DJIA, utilizzando semplicemente la sua serie storica pubblicamente disponibile. Per migliorare l'affidabilità dei risultati rispetto alla letteratura, abbiamo sperimentato l'approccio su un lungo periodo consecutivo di 18 anni </ins>di <ins style="font-weight: bold; text-decoration: none;">serie storiche DJIA, dal 2000 al 2017. In 8 anni di trading nel periodo del set di test dal 2009 al 2017, la soluzione ha quintuplicato il capitale iniziale, inoltre poiché il DJIA ha in media un trend crescente, abbiamo testato l'approccio anche con un trend mediamente decrescente invertendo semplicemente la stessa serie storica del DJIA. In questo caso estremo, in cui difficilmente un investitore rischierebbe denaro, l'approccio ha più che raddoppiato il capitale iniziale.</ins></div></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div><del style="font-weight: bold; text-decoration: none;">Topic: [[Topic::Efficienza </del>delle RNN <del style="font-weight: bold; text-decoration: none;">nella generazione </del>di <del style="font-weight: bold; text-decoration: none;">testi]]</del></div></td><td colspan="2" class="diff-side-added"></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div> </div></td><td colspan="2" class="diff-side-added"></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div><del style="font-weight: bold; text-decoration: none;">__SHOWFACTBOX__</del></div></td><td colspan="2" class="diff-side-added"></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div> </div></td><td colspan="2" class="diff-side-added"></td></tr> <tr><td class="diff-marker" data-marker="−"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #ffe49c; vertical-align: top; white-space: pre-wrap;"><div><del style="font-weight: bold; text-decoration: none;">[[Category:pubblicazione]]</del></div></td><td colspan="2" class="diff-side-added"></td></tr> </table>

Mindmakerbot https://wiki.mindmaker.it/index.php?title=The_Unreasonable_Effectiveness_of_Recurrent_Neural_Networks&diff=1873&oldid=prev 2024-04-09T10:44:54Z

<p></p> <table style="background-color: #fff; color: #202122;" data-mw="interface"> <col class="diff-marker" /> <col class="diff-content" /> <col class="diff-marker" /> <col class="diff-content" /> <tr class="diff-title" lang="it"> <td colspan="2" style="background-color: #fff; color: #202122; text-align: center;">← Versione meno recente</td> <td colspan="2" style="background-color: #fff; color: #202122; text-align: center;">Versione delle 10:44, 9 apr 2024</td> </tr><tr><td colspan="2" class="diff-lineno" id="mw-diff-left-l10">Riga 10:</td> <td colspan="2" class="diff-lineno">Riga 10:</td></tr> <tr><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><br></td><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><br></td></tr> <tr><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>__SHOWFACTBOX__</div></td><td class="diff-marker"></td><td style="background-color: #f8f9fa; color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #eaecf0; vertical-align: top; white-space: pre-wrap;"><div>__SHOWFACTBOX__</div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;"></ins></div></td></tr> <tr><td colspan="2" class="diff-side-deleted"></td><td class="diff-marker" data-marker="+"></td><td style="color: #202122; font-size: 88%; border-style: solid; border-width: 1px 1px 1px 4px; border-radius: 0.33em; border-color: #a3d3ff; vertical-align: top; white-space: pre-wrap;"><div><ins style="font-weight: bold; text-decoration: none;">[[Category:pubblicazione]]</ins></div></td></tr> </table>

Alesaccoia https://wiki.mindmaker.it/index.php?title=The_Unreasonable_Effectiveness_of_Recurrent_Neural_Networks&diff=1859&oldid=prev 2024-04-09T09:21:07Z

<p>Creata pagina con &quot;Titolo: <a href="/index.php?title=Titolo::The_Unreasonable_Effectiveness_of_Recurrent_Neural_Networks&amp;action=edit&amp;redlink=1" class="new" title="Titolo::The Unreasonable Effectiveness of Recurrent Neural Networks (la pagina non esiste)">Titolo::The Unreasonable Effectiveness of Recurrent Neural Networks</a> Anno di pubblicazione: <a href="/index.php?title=AnnoDiPubblicazione::2015&amp;action=edit&amp;redlink=1" class="new" title="AnnoDiPubblicazione::2015 (la pagina non esiste)">AnnoDiPubblicazione::2015</a> Autore: <a href="/index.php?title=Autore::Andrej_Karpathy&amp;action=edit&amp;redlink=1" class="new" title="Autore::Andrej Karpathy (la pagina non esiste)">Autore::Andrej Karpathy</a> URL: <a href="/index.php?title=URL::https://karpathy.github.io/2015/05/21/rnn-effectiveness/&amp;action=edit&amp;redlink=1" class="new" title="URL::https://karpathy.github.io/2015/05/21/rnn-effectiveness/ (la pagina non esiste)">URL::https://karpathy.github.io/2015/05/21/rnn-effectiveness/</a> Topic: <a href="/index.php?title=Topic::Efficienza_delle_RNN_nella_generazione_di_testi&amp;action=edit&amp;redlink=1" class="new" title="Topic::Efficienza delle RNN nella generazione di testi (la pagina non esiste)">Topic::Efficienza delle RNN nella generazione di testi</a> __SHOWFACTBOX__&quot;</p> <p><b>Nuova pagina</b></p><div>Titolo: [[Titolo::The Unreasonable Effectiveness of Recurrent Neural Networks]]<br /> <br /> Anno di pubblicazione: [[AnnoDiPubblicazione::2015]]<br /> <br /> Autore: [[Autore::Andrej Karpathy]]<br /> <br /> URL: [[URL::https://karpathy.github.io/2015/05/21/rnn-effectiveness/]]<br /> <br /> Topic: [[Topic::Efficienza delle RNN nella generazione di testi]]<br /> <br /> __SHOWFACTBOX__</div>

Sara Maserati