Iskalni niz:
išči po
išči po
išči po
išči po
Vrsta gradiva:
Jezik:
Št. zadetkov: 7
Sistem za avtomatsko ocenjevanje algoritmov
Martin Jakomin, Tomaž Dobravec
Diplomsko delo
Oznake: algoritmi;ocenjevanje kakovosti algoritmov;avtomatsko ocenjevanje kakovosti algoritmov;spletni sistem;spletni vmesnik;računalništvo;računalništvo in informatika;računalništvo in matematika;univerzitetni študij;interdisciplinarni študij;diplomske naloge;
Sistem za avtomatsko ocenjevanje algoritmov
Leto: 2012 Vir: Fakulteta za računalništvo in informatiko (UL FRI)
Incremental matrix factorization for simultaneous learning from parallel data streams
Martin Jakomin, Zoran Bosnić, Tomaž Curk
Doktorska disertacija
Oznake: machine learning;matrix factorization;data streams;data fusion;incremental learning;recommender systems;synthetic data generator;computer and information science;doctoral dissertations;
Matrix factorization techniques have proven to be useful and reliable for solving largescale machine learning problems. The data sparsity and cold-start problems found in real-world applications, such as recommender systems, can be indirectly alleviated by considering multiple heterogeneous data ...
Leto: 2019 Vir: Fakulteta za računalništvo in informatiko (UL FRI)
Optimizacija nakupovanja spletnega oglasnega prostora
Jakob Bevc, Jure Demšar, Martin Jakomin
Magistrsko delo
Oznake: spletno oglaševanje;dražbe v realnem času;napoved porazdelitve ponudb;računalništvo;računalništvo in informatika;magisteriji;
V procesu nakupovanja spletnega oglasnega prostora dandanes prevladuje tako imenovano programatično oglaševanje. Gre za trženje oglasnega prostora preko avtomatiziranih dražb v realnem času. Za uspešno sodelovanje na dražbah je potrebno usklajeno delovanje različnih procesov, v katerih ključno vlogo ...
Leto: 2021 Vir: Fakulteta za računalništvo in informatiko (UL FRI)
Optimizing click-trough rate in online advertising using cost-sensitive learning
Bernarda Petek, Zoran Bosnić, Martin Jakomin
Magistrsko delo
Oznake: real-time bidding;click-trough rate prediction;cost-sensitive learning;custom loss function;
Real-time bidding is a type of online advertising, which displays personalized advertisements online to users based on their interests in real time. Demand-side platforms participate in such bidding for ad spaces. The bidding price, usually computed using a predicted click-through rate and target co ...
Leto: 2023 Vir: Fakulteta za računalništvo in informatiko (UL FRI)
Optimizing real-time bidding systems in programmatic advertising
Marko Možina, Jure Demšar, Martin Jakomin
Magistrsko delo
Oznake: programmatic advertising;real-time bidding;candidate retrieval;two-tower models;sampled softmax;mixed negative sampling;sampling bias correction;online A/B testing;computer science;master's thesis;
A demand-side platform has only tens of milliseconds per request to choose the advertisement it will bid with, so the selection runs in stages. The first stage carries much of that weight. A computationally cheap two-tower model scores the large set of eligible candidates and forwards about a hu ...
Leto: 2026 Vir: Fakulteta za računalništvo in informatiko (UL FRI)
Multi-task learning in programmatic advertising
Domen Vreš, Marko Robnik Šikonja, Martin Jakomin
Magistrsko delo
Oznake: artificial intelligence;machine learning;multi–task learning;real–time bidding;viewability prediction;click–through rate prediction;conversion rate prediction;deep & cross network;cross–stitch layer;soft attention;uncertainty based loss weighing;computer science;master's thesis;
With the popularity of the world wide web, online advertising became crucial for the advertising industry. A large part of online advertising is based on real-time bidding, where automated software is used to buy and sell ad space. In this ecosystem, demand-side platforms (DSP) work with advertisers ...
Leto: 2023 Vir: Fakulteta za računalništvo in informatiko (UL FRI)
Utility of embeddings in multi-modal models for click-through rate prediction
Mark Žnidar, Marko Robnik Šikonja, Blaž Škrlj, Martin Jakomin
Diplomsko delo
Oznake: machine learning;click-through rate prediction;programmatic advertising;multi-modal embeddings;real-time bidding;computer science;diploma;
Accurate click-through rate (CTR) prediction under strict real-time bidding latency constraints is central to modern programmatic advertising. Models powering large-scale recommender systems rely on high-cardinality categorical IDs and neglect the rich semantic cues present in ad creatives and page ...
Leto: 2025 Vir: Fakulteta za računalništvo in informatiko (UL FRI)
Št. zadetkov: 7
Ključne besede:
Leto izdaje:
Avtorji:
Repozitorij:
Tipologija:
Jezik: