{"id":5936,"date":"2026-10-15T15:41:26","date_gmt":"2026-10-15T15:41:26","guid":{"rendered":"https:\/\/cog-ist.com\/?post_type=duyuru&p=5936"},"modified":"2026-10-15T15:44:53","modified_gmt":"2026-10-15T15:44:53","slug":"psikoloji-ve-norobilimde-makine-ogrenmesine-giris-dr-yusuf-can-semerci-egitim-57","status":"publish","type":"duyuru","link":"https:\/\/cog-ist.com\/en\/duyuru\/psikoloji-ve-norobilimde-makine-ogrenmesine-giris-dr-yusuf-can-semerci-egitim-57\/","title":{"rendered":"Psikoloji ve N\u00f6robilimde Makine \u00d6\u011frenmesine Giri\u015f – Dr. Yusuf Can Semerci | E\u011fitim #57"},"content":{"rendered":"

Psikoloji <\/strong>ve neuroscience <\/strong>alanlar\u0131nda yapay zekay\u0131 <\/strong>farkl\u0131 \u015fekillerde kullan\u0131yoruz. Bunlar\u0131n ba\u015f\u0131nda \u00e7e\u015fitli altba\u015fl\u0131klar\u0131yla makine \u00f6\u011frenmesi <\/strong>geliyor. Psikoloji <\/strong>ve n\u00f6robilimde makine \u00f6\u011frenmesini <\/strong>nas\u0131l kullan\u0131yoruz<\/strong>? Yapay zekadan <\/strong>kendi ara\u015ft\u0131rmalar\u0131m\u0131zda<\/strong>, kendi bilimsel \u00e7al\u0131\u015fmalar\u0131m\u0131zda <\/strong>nas\u0131l faydalanabiliriz<\/strong>? Bunun i\u00e7in \u00f6\u011frenmeye nereden ba\u015flamak gerekir?<\/strong><\/p>\n\n\n\n

\u00c7o\u011fu ki\u015fi bu konular\u0131 \u00f6\u011frenmek <\/strong>istedi\u011finde sonsuz teknik detay <\/strong>ve matematiksel g\u00f6sterim <\/strong>dolay\u0131s\u0131yla biraz \u00fcrker<\/strong>. Motivasyonunu koruyamaz… Bizce, teknik <\/strong>ve matematiksel detaylar\u0131 <\/strong>bir kenara ay\u0131r\u0131p <\/strong>\u00f6nce kavramsal temelleri \u00f6\u011frenmek herkes i\u00e7in en sa\u011fl\u0131kl\u0131 yol! <\/strong>Dolay\u0131s\u0131yla makine \u00f6\u011frenmesine<\/strong>, daha spesifik olarak makine <\/strong>\u00f6\u011frenmesinin psychology <\/strong>ve n\u00f6robilimdeki <\/strong>kullan\u0131m\u0131na kavramsal <\/strong>ve ku\u015fbak\u0131\u015f\u0131 <\/strong>bir giri\u015f yap\u0131yoruz. <\/p>\n\n\n\n

En ba\u015fta bu alanlarda \u00f6\u011frenci veya ara\u015ft\u0131rmac\u0131lar, ard\u0131ndan ilgili herkes alan\u0131nda uzman Dr. Yusuf Can Semerci rehberli\u011finde konuya kavramsal ve kapsaml\u0131 bir giri\u015f yapabilir!<\/strong><\/p>\n\n\n\n

Ders Tan\u0131t\u0131m\u0131:<\/p>\n\n\n\n

Makine \u00f6\u011frenmesi<\/strong>, verideki \u00f6r\u00fcnt\u00fcleri ke\u015ffetmemizi <\/strong>sa\u011flayan hesaplamal\u0131 (computational, i\u015flemlemesel)<\/strong> bir yakla\u015f\u0131m. Beyin g\u00f6r\u00fcnt\u00fcleme, EEG <\/strong>ve davran\u0131\u015f <\/strong>verilerinin h\u0131zla b\u00fcy\u00fcd\u00fc\u011f\u00fc g\u00fcn\u00fcm\u00fczde bu y\u00f6ntemleri anlamak, cognitive science <\/strong>ile ilgilenen herkes i\u00e7in vazge\u00e7ilmez hale geliyor. Bir t\u00fcm\u00f6r\u00fc MRI g\u00f6r\u00fcnt\u00fcs\u00fcnden ay\u0131rt etmekten EEG sinyallerinde zamansal \u00f6r\u00fcnt\u00fcler bulmaya kadar, makine \u00f6\u011frenmesi<\/strong>nin hik\u00e2yesi ayn\u0131 zamanda insan beyninin \u00f6\u011frenme bi\u00e7imlerinden ilham alan algoritmalar\u0131n hik\u00e2yesidir.<\/p>\n\n\n\n

D\u00f6rt hafta s\u00fcrecek bu e\u011fitim program\u0131 farkl\u0131 disiplinlerden kat\u0131l\u0131mc\u0131lar i\u00e7in tasarland\u0131. Matematik ya da programlama altyap\u0131s\u0131 gerektirmiyor. <\/strong>Kat\u0131l\u0131mc\u0131lar verinin <\/strong>ne oldu\u011fundan ve nas\u0131l temsil edildi\u011finden <\/strong>ba\u015flayarak k\u00fcmeleme<\/strong>, denetimli \u00f6\u011frenme <\/strong>ve sinir a\u011flar\u0131na <\/strong>uzanan bir yolculukta temel algoritmalar\u0131n <\/strong>nas\u0131l \u00e7al\u0131\u015ft\u0131\u011f\u0131n\u0131, nas\u0131l de\u011ferlendirildi\u011fini ve verilere nas\u0131l uyguland\u0131\u011f\u0131n\u0131 \u00f6\u011frenecekler.<\/p>\n\n\n\n

E\u011fitmenimiz Dr. Yusuf Can Semerci<\/strong>, Maastricht \u00dcniversitesi <\/strong>Fen ve M\u00fchendislik Fak\u00fcltesi \u0130leri Hesaplama Bilimleri <\/strong>(Advanced Computing Sciences) b\u00f6l\u00fcm\u00fcnde Yard\u0131mc\u0131 Do\u00e7ent <\/strong>olarak g\u00f6rev yapmaktad\u0131r. Doktoras\u0131n\u0131 Yeditepe \u00dcniversitesi Bilgisayar M\u00fchendisli\u011fi<\/strong> b\u00f6l\u00fcm\u00fcnde <\/strong>tamamlam\u0131\u015f, 2011-2020 y\u0131llar\u0131 aras\u0131nda ayn\u0131 \u00fcniversitede ara\u015ft\u0131rma g\u00f6revlisi olarak \u00e7al\u0131\u015fm\u0131\u015ft\u0131r. Bu d\u00f6nemde ayr\u0131ca Siemens T\u00fcrkiye<\/strong> (2014-2016) ve Cenasis Software<\/strong>‘de (2016-2018) yaz\u0131l\u0131m geli\u015ftirme alan\u0131nda deneyim kazanm\u0131\u015f, 2019-2020’de Bo\u011fazi\u00e7i \u00dcniversitesi’<\/strong>nde misafir ara\u015ft\u0131rmac\u0131 <\/strong>olarak bulunmu\u015ftur. Doktora s\u00fcrecinde Timeline Travel ve EMBOA (Erasmus+) projelerinde ara\u015ft\u0131rmac\u0131 olarak yer alm\u0131\u015f, Ocak 2021’de Maastricht \u00dcniversitesi’ne doktora sonras\u0131 ara\u015ft\u0131rmac\u0131 <\/strong>olarak kat\u0131lm\u0131\u015f ve Ocak 2025’te yard\u0131mc\u0131 do\u00e7ent <\/strong>olmu\u015ftur.<\/p>\n\n\n\n

\u00c7al\u0131\u015fmalar\u0131nda insan davran\u0131\u015f\u0131 ve insan-bilgisayar etkile\u015fimi<\/strong> \u00fczerine yo\u011funla\u015f\u0131yor, bu alanlarda yapay zek\u00e2 temelli hesaplamal\u0131<\/strong> y\u00f6ntemler kullan\u0131yor. \u0130lgi alanlar\u0131 aras\u0131nda duygusal hesaplama <\/strong>(affective computing), e\u011fitsel bili\u015fim <\/strong>(educational computing), do\u011fal dil i\u015fleme<\/strong> (natural language processing) ve bilgisayarl\u0131 g\u00f6r\u00fc <\/strong>(computer vision) yer al\u0131yor. \u0130nsan Merkezli Yapay Zek\u00e2 Ara\u015ft\u0131rma Grubu<\/strong>‘nun e\u015f-koordinat\u00f6rl\u00fc\u011f\u00fcn\u00fc y\u00fcr\u00fcten ve Bili\u015fsel Sistemler Ara\u015ft\u0131rma Grubu<\/strong>‘nun (Cognitive Systems Research Group) \u00fcyesi olan Dr. Semerci, VOXReality, XR2Learn, JOINclusion ve PROCare4Life gibi disiplinler aras\u0131 projelerde yer alm\u0131\u015ft\u0131r.<\/p>\n\n\n\n

E\u011fitim 5, 12, 19, 26 Kas\u0131m Per\u015fembe g\u00fcnleri saat 19:30 \u2013 21:30 aras\u0131nda \u00e7evrimi\u00e7i olarak Zoom uygulamas\u0131 \u00fczerinden ger\u00e7ekle\u015fecek. Toplamda yakla\u015f\u0131k 8 saat s\u00fcrecek. T\u00fcm oturumlar kaydedilmekte ve kat\u0131l\u0131mc\u0131larla payla\u015f\u0131lmaktad\u0131r. Oturumlara e\u011fitim bittikten sonra da eri\u015filebilir.<\/strong><\/p>\n\n\n\n

<\/div>\n\n\n\n
\n\n\n\n
<\/div>\n\n\n\n

1. Hafta (5 Kas\u0131m): Temeller – Veri, \u00d6znitelikler ve \u00d6\u011frenme<\/strong><\/p>\n\n\n\n

<\/p>\n\n\n\n

    \n
  • Makine \u00f6\u011frenmesi nedir?<\/li>\n\n\n\n
  • Veri t\u00fcrleri: tablo, g\u00f6r\u00fcnt\u00fc, video, sinyal ve metin<\/li>\n\n\n\n
  • Beyin bilimi verileri: EEG, MEG, MRI\/fMRI ve davran\u0131\u015fsal veri<\/li>\n\n\n\n
  • De\u011fi\u015fken t\u00fcrleri ve \u00f6znitelikler (features)<\/li>\n\n\n\n
  • \u00d6\u011frenme paradigmalar\u0131: denetimli, denetimsiz ve peki\u015ftirmeli \u00f6\u011frenme<\/li>\n\n\n\n
  • Model de\u011ferlendirme: a\u015f\u0131r\u0131 \u00f6\u011frenme (overfitting), yetersiz \u00f6\u011frenme (underfitting), e\u011fitim\/do\u011frulama\/test ayr\u0131m\u0131<\/li>\n<\/ul>\n\n\n\n

    <\/p>\n\n\n\n

    Bu oturumda makine \u00f6\u011frenmesinin temel mant\u0131\u011f\u0131n\u0131, hangi t\u00fcr verilerle \u00e7al\u0131\u015ft\u0131\u011f\u0131n\u0131 ve bir modelin iyi \u00f6\u011frendi\u011fini nas\u0131l anlayaca\u011f\u0131m\u0131z\u0131 konu\u015faca\u011f\u0131z.<\/p>\n\n\n\n

    <\/p>\n\n\n\n

    2. Hafta (12 Kas\u0131m): Denetimsiz \u00d6\u011frenme \u2013 K\u00fcmeleme<\/strong><\/p>\n\n\n\n

    <\/p>\n\n\n\n

      \n
    • Denetimsiz \u00f6\u011frenme: etiketsiz veride gizli yap\u0131y\u0131 ke\u015ffetmek<\/li>\n\n\n\n
    • K\u00fcmeleme y\u00f6ntemleri: merkez, ba\u011flant\u0131, yo\u011funluk ve da\u011f\u0131l\u0131m temelli yakla\u015f\u0131mlar<\/li>\n\n\n\n
    • K-Means algoritmas\u0131 ve ba\u015flang\u0131\u00e7 y\u00f6ntemleri (naive sharding)<\/li>\n\n\n\n
    • K\u00fcmeleme sonu\u00e7lar\u0131n\u0131 de\u011ferlendirmek: sahte \u00f6r\u00fcnt\u00fcler ve yorumlama zorluklar\u0131<\/li>\n\n\n\n
    • Boyut indirgemeye giri\u015f: PCA, t-SNE<\/li>\n<\/ul>\n\n\n\n

      <\/p>\n\n\n\n

      Bu oturumda verinin i\u00e7indeki gruplar\u0131 nas\u0131l bulabilece\u011fimizi inceleyece\u011fiz.<\/p>\n\n\n\n

      <\/p>\n\n\n\n

      3. Hafta (19 Kas\u0131m): Denetimli \u00d6\u011frenme – Regresyon ve S\u0131n\u0131fland\u0131rma<\/strong><\/p>\n\n\n\n

      <\/p>\n\n\n\n

        \n
      • Etiketli veriden \u00f6\u011frenmek: girdi-\u00e7\u0131kt\u0131 ili\u015fkisi ve etiketleme<\/li>\n\n\n\n
      • Regresyon ve s\u0131n\u0131fland\u0131rma: s\u00fcrekli de\u011ferler ve ayr\u0131k kategoriler<\/li>\n\n\n\n
      • E\u011fitim hedefi: a\u011f\u0131rl\u0131klar\u0131 g\u00fcncellemek, hata fonksiyonlar\u0131 (MSE, \u00e7apraz entropi)<\/li>\n\n\n\n
      • Do\u011frusal regresyon: n\u00f6ron ate\u015fleme h\u0131z\u0131 ve tepki s\u00fcresi \u00f6rne\u011fi<\/li>\n\n\n\n
      • Destek Vekt\u00f6r Makineleri (SVM): marjin, e\u015fik ve yumu\u015fak marjin<\/li>\n<\/ul>\n\n\n\n

        <\/p>\n\n\n\n

        Bu oturumda en iyi \u00e7izgiyi \u00e7izmek ile veri noktalar\u0131n\u0131 do\u011fru kutuya koymak aras\u0131ndaki fark\u0131 \u00f6\u011frenece\u011fiz.<\/p>\n\n\n\n

        <\/p>\n\n\n\n

        4. Hafta (26 Kas\u0131m): \u0130leri Modeller – Sinir A\u011flar\u0131ndan Derin \u00d6\u011frenmeye<\/strong><\/p>\n\n\n\n

        <\/p>\n\n\n\n

          \n
        • Yapay n\u00f6ron: biyolojik n\u00f6rondan ilham, a\u011f\u0131rl\u0131kl\u0131 toplam ve aktivasyon fonksiyonlar\u0131<\/li>\n\n\n\n
        • Perseptron, yapay sinir a\u011flar\u0131 ve gradyan ini\u015fi<\/li>\n\n\n\n
        • Derin \u00f6\u011frenme: neden derin a\u011flara ihtiya\u00e7 duyduk? (Otomatik \u00f6znitelik \u00e7\u0131kar\u0131m\u0131, gradyan kaybolmas\u0131 sorunu)<\/li>\n\n\n\n
        • Evri\u015fimli Sinir A\u011flar\u0131 (CNN): g\u00f6rsel korteksten ilham alan yap\u0131lar<\/li>\n\n\n\n
        • Yinelemeli Sinir A\u011flar\u0131 (RNN) ve LSTM: zamansal sinyaller i\u00e7in yap\u0131lar<\/li>\n\n\n\n
        • Otokodlay\u0131c\u0131lar, GAN’lar ve Transformer: modern mimarilere genel bak\u0131\u015f<\/li>\n<\/ul>\n\n\n\n

          <\/p>\n\n\n\n

          Son oturumda yapay sinir a\u011flar\u0131n\u0131n beyinden nas\u0131l esinlendi\u011fini konu\u015facak, farkl\u0131 veri t\u00fcrleri i\u00e7in hangi derin mimarinin uygun oldu\u011funu inceleyece\u011fiz.<\/p>\n\n\n\n

          <\/div>\n\n\n\n
          \n\n\n\n
          <\/div>\n\n\n\n

          Etkinli\u011fin \u00fccreti ki\u015fi ba\u015f\u0131 KDV dahil 1850 TL\u2019dir.<\/strong><\/p>\n\n\n\n

          \u00d6deme bilgileri, kay\u0131t formunu dolduran kat\u0131l\u0131mc\u0131lar ile en k\u0131sa zamanda payla\u015f\u0131lacakt\u0131r.<\/p>\n\n\n\n

          <\/div>\n\n\n\n

          <\/p>","protected":false},"featured_media":5938,"parent":0,"template":"","meta":{"_acf_changed":false},"kategori":[124,40,41,42],"class_list":["post-5936","duyuru","type-duyuru","status-publish","has-post-thumbnail","hentry","kategori-bilissel-bilim","kategori-norobilim","kategori-psikoloji","kategori-yapayzeka"],"acf":[],"_links":{"self":[{"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/duyuru\/5936","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/duyuru"}],"about":[{"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/types\/duyuru"}],"version-history":[{"count":3,"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/duyuru\/5936\/revisions"}],"predecessor-version":[{"id":5941,"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/duyuru\/5936\/revisions\/5941"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/media\/5938"}],"wp:attachment":[{"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/media?parent=5936"}],"wp:term":[{"taxonomy":"kategori","embeddable":true,"href":"https:\/\/cog-ist.com\/en\/wp-json\/wp\/v2\/kategori?post=5936"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}