We introduce a model considering the single-leg revenue management problem from the perspective of online algorithms and competitive analysis. In this model, the price and limitation of bookings are both decision vari...
详细信息
Budget-related decisions in search auctions are recognized as a structured decision problem, rather than a simple constraint. Budget planning over several coupled campaigns remains a challenging but utterly important ...
详细信息
Budget-related decisions in search auctions are recognized as a structured decision problem, rather than a simple constraint. Budget planning over several coupled campaigns remains a challenging but utterly important task in search advertisements. In this paper, we propose a multi-campaign budget plan- ning approach using optimal control techniques, with consideration of the substitute relationship between advertising campaigns. A measure of coupled relationships between campaigns is presented, e.g., the over- lapping degree (O) in terms of campaign contents, promotional periods and target regions. We also discuss some desirable properties of our model and possible solutions. Furthermore, computational experiments are conducted to evaluate our model and identified properties, with real-world data collected from logs and reports of practical campaigns. Experimental results show that, (a) coupled campaigns with higher over- lapping degrees can reduce the optimal budget level and the optimal revenue, and also arrive the budgeting cap earlier;(b) The advertising effort could be seriously weakened when ignoring the overlapping degree between campaigns.
From the view of combination of stereo vision and robot control, an active stereo vision for object tracking and location system is built to work on flexible manufacturing system. A method based on local HSV image and...
详细信息
From the view of combination of stereo vision and robot control, an active stereo vision for object tracking and location system is built to work on flexible manufacturing system. A method based on local HSV image and shape of object to recognize object is used for robot tracking. After the color segment, use the knowledge of shape of objects to recognize objects. And a technique for camera calibration based on plane square is proposed. It only requires the camera to observe a square at two (or more) different unknown orientations. Use the result of object recognize and camera calibration the robot can position and track the object. At the last of the paper, the results of object tracking experiments prove that the method is correct.
An important characteristic of real-world learning process is that the data frequently contains uncertainties. The uncertainties in the datasets deteriorate the learning process. Hence, to properly represent and handl...
详细信息
One of challenges in GIS Web service is mass GIS data storage and processing. There are at least three obstacles that oppose dealing with this challenge: the mismatch between the frequent requirement of concurrent tas...
详细信息
One of challenges in GIS Web service is mass GIS data storage and processing. There are at least three obstacles that oppose dealing with this challenge: the mismatch between the frequent requirement of concurrent tasks for accessing data in GIS Web service and the poor performance of conventional storage mechanism such as relational databases;the insufficiency of relational databases to manage complex data types in GIS Web service;the horizontal scalab.lity of storage for non-stop GIS Web service evolution. This paper proposes a distributed cached storage solution, as the mixture of MongoDB storage and Memcached cache, for GIS Web service. Both techniques are closely related to NoSQL that is a class of database management system identified by its non-adherence to the widely used relational database management system (RDBMS) model. A concise case that facilitates a better understanding of this work indicates its feasibility.
Important users are high-status vertices in social networks. They are everywhere in most fields of society and have big impact on those around them. Although a lot of effort has been made on identifying important user...
详细信息
ISBN:
(纸本)9781467330930
Important users are high-status vertices in social networks. They are everywhere in most fields of society and have big impact on those around them. Although a lot of effort has been made on identifying important users, the efficient methods still need to be developed, especially for the web users from Sina microblog, which is the most popular social networking sites in China and has unique characteristics. In this paper, a machine learning-based method which only uses several attributes on Naive Bayes Classifiers (NBC) and Back Propagation Neural Network (BPNN) was proposed to identify important users. Initial experiments indicate that our method is effective. The result of "high" category has more than 55% accuracy rate. We find the NBC can identify more important users while BPNN has higher accuracy rate. What's more, the numbers of follower and followings in Sina microblog is independent.
Radio fingerprint matching localization method promises high localization accuracy but requires extensive infrastructural effort and search operations. This paper proposes a new search strategy for radio fingerprint m...
详细信息
In this paper, we study the budget-constrained bidding problems in sponsored search. Our findings illustrate that, compared to budget-irrespective bidding, advertisers can obtain equal even better return-on-investment...
详细信息
In this paper, we study the budget-constrained bidding problems in sponsored search. Our findings illustrate that, compared to budget-irrespective bidding, advertisers can obtain equal even better return-on-investment (ROI) when considering theirs and the competitor's budget constrains.
This paper proposes a novel budget model based on differential game to deal with budget allocation in competitive search advertisements under a finite time horizon, with consideration of budget constraints. We extend ...
详细信息
This paper proposes a novel budget model based on differential game to deal with budget allocation in competitive search advertisements under a finite time horizon, with consideration of budget constraints. We extend the advertising response function with the dynamical advertising effort u and quality score q to fit search advertising scenarios. We also discuss Nash equilibriums of our model, and study some desirable properties of two kinds of equilibriums in the case with budget constraints: "budget-stable" open-loop Nash equilibrium (BS-OLNE) and "budget-unstable" open-loop Nash equilibrium (BUS-OLNE). We have evaluated our budget model and identified properties with computational experiments. Experimental results show that budget strategies with dynamical advertising elasticity are superior to those with fixed one and our findings on OLNEs are helpful for advertisers to make budget decisions.
Polarity shifting has been a challenge to automatic sentiment classification. In this paper, we create a corpus which consists of polarity-shifted sentences in various kinds of product reviews. In the corpus, both the...
详细信息
暂无评论