基于LASSO变量选择方法的网络广告点击率预测模型研究

与传统的的媒体营销模式相比,搜索引擎广告因其精准和投入低等特点获得巨大成功。但已有的搜索引擎广告点击率模型不能有效解决数据量大及特征维度高的问题,使预测结果的准确性大打折扣。本文构建了一种基于LASSO变量选择方法的广告点击率预测模型,能有效克服现有广告点击率模型在处理数据高维性和稀疏性方面的不足。利用某公司的竞价数据对模型进行验证,结果表明影响广告点击率的关键因素是广告关键词中的商标信息、地域信息和每点击成本。该研究结果为企业制定搜索引擎广告营销策略提供一定的理论依据。     

带赔偿的云计算服务拍卖机制研究

云计算提供了按需获取、按使用付费的IT资源获取模式,由于存在大量的云计算服务提供商,以市场化的形式来组织云计算资源的分配与获取是一种合理的方式.以面向市场的云计算服务架构为基础,研究了云计算用户和服务提供商之间的匹配问题.采用拍卖方式实现具体的匹配过程,并将服务失败时的赔偿问题纳入考虑,结合服务提供商的违约率提出了两种带赔偿的云计算服务拍卖机制一在逆向一级密封拍卖上增加赔偿因子,分析了其二人博弈和多人博弈的均衡;同时,研究了服务权再分配对该机制的影响.     

固定投资回报型合同节水管理项目收益分配模型

合同节水管理(WSMC)是一种用节约的水费支付改造成本并获取收益的节水模式,主要涉及用户与节水服务公司两个参与方。本文研究用户无法观测节水服务公司行动条件下固定投资回报型WSMC中的收益分配问题,该模式优先偿还投资,其次分享收益的特征需要激励机制诱使节水服务公司增加节水收益。首先,选取改造成本和节水量作为激励因素,设计用户对节水服务公司的激励合同,并构建节水量产出函数。其次,以用户的期望效用最大化为目标,建立用户对节水服务公司的激励模型并采用逆推法求解模型。再次,对均衡结果进行分析,结果表明:节水服务公司的最优成本分享比例和最优节水量分享比例与风险规避程度、努力成本系数和项目不确定性负相关;此外,最优节水量分享比例与综合能力正相关。最后,通过数值分析探讨分享比例和努力程度随合同参数的变化情况。     

网络视频节目中羊群效应的影响研究

考虑累积观看节目用户量产生的羊群效应角度构建两阶段决策模型,得到节目最优定价策略和最优广告量策略。研究表明:收费模式中羊群效应对最优定价的影响特征与用户规模有关,并进一步导致免费模式中最优广告量减少。在收费模式羊群效应较高情况下,运营商能通过延长收费模式时间提高节目利润。还发现收费模式中羊群效应导致观看节目用户总规模增加,但收费模式对免费模式有挤兑效应。在免费模式中羊群效应较低情况下,只有收费模式中具有较高的羊群效应才能使节目总利润增加,否则总利润减少。     

国债招标拍卖的最优机制:数量与价格歧视

过去已有相当多的文献讨论国债拍卖,米勒和弗雷德曼认为统一价格比歧视价格拍卖有比较优势.研究国债拍卖最优分配方式的选择问题,采用最优机制设计方法分析国债拍卖活动,内生地得到最优拍卖机制.然而结果表明,如果使用价格歧视及数量歧视(定量分配),那么政府的销售收益会得到改善,这是一个最优拍卖机制,不同于米勒和弗雷德曼的观点.     

考虑消费者策略行为的全渠道两阶段广告投放与动态定价决策

在需求为导向的全渠道市场环境下，切实从消费者行为属性制定广告与价格营销策略是促进全渠道零售长足发展的关键。本文将消费者策略行为引入到全渠道零售中，构建零售商在两阶段均不投放广告、第一阶段正常期投放广告及第二阶段优惠期投放广告等三种广告决策模式下全渠道两阶段广告投放与动态定价模型，探讨全渠道零售商的最优响应策略，并对模型进行数值算例。研究发现：当消费者的策略性水平较低时，零售商选择在正常期(优惠期)投放广告策略下，正常期产品定价较高(较低),优惠期产品定价较低(较高);零售商在正常期或优惠期投放广告对正常期销量都是有利的；当消费者的策略性水平较低或广告的影响系数较低时，零售商选择在正常期投放广告能获得最优收益；当消费者的策略性水平偏高、广告的影响系数适中时，零售商在优惠期投放广告是最优的；当消费者的策略性水平及广告的影响系数都偏高时，零售商选择在两阶段均不投放广告是最优策略。     

两期环境下闭环供应链的销售渠道选择策略

针对DI(电子渠道直销新产品及通过零售商分销再制造产品)、ID(分销新产品与直销再制造产品)和DD(同时直销新产品与再制造产品)的三种销售模式,分别建立两周期闭环供应链博弈(优化)模型探讨了成员的最优定价策略、回收率与利润分配问题,并通过理论推导和数值算例将电子渠道中的消费者附加成本对三种模式均衡解与利润的影响进行了分析,研究结果表明:1)当电子渠道的附加成本为零时,DD模式是制造商与系统的最优选择,其次是DI模式,最后是ID模式;而从零售商与环境绩效的角度则是ID模式最优.2)当电子渠道的附加成本非零时,三种模式均可能成为制造商与闭环供应链系统的最优选择,但DI模式永不会成为最劣选择;三种模式下零售商收益均随附加成本的增加而得到改善,且零售商仍愿意选择ID模式.     

基于货到付款支付模式且考虑银行存贷的二级供应链Stackelberg定价决策

在货到付款支付模式下二级供应链定价决策中,供应链企业资金闲置时向银行存款或资金约束时向银行贷款(银行存贷)的行为是不可忽视的重要因素,如何构建基于货到付款支付模式且考虑银行存贷的二级供应链Stackelberg定价决策模型是需要关注的重要问题。在本文中,首先给出了市场需求函数;然后,基于货到付款支付模式,针对制造商资金或零售商资金约束情形,分别构建针对不同供应链权力结构的定价决策模型;进一步地,通过模型求解确定了不同情形下不同权力结构的制造商与零售商的最优策略,并分析了模型参数对最优策略的影响;最后,针对不同资金约束情形与不同权力结构的最优策略以及银行利率对最优策略及利润影响,给出了对比分析。研究表明三种银行利率均会影响最优策略,且资金约束对象差异的影响明显。     

新能源背景下考虑政府补贴的汽车供应链决策研究

研究了政府对新能源汽车实施政府补贴政策所采取的两种模式:直接补贴制造商(GM模式)与直接补贴销售商(GS模式),并建立了两种不同政府补贴模式的汽车供应链模型.通过比较不同模式下汽车供应链各方及政府补贴政策的最优决策,发现制造商作为市场的主导者,相比销售商来说具有更强的讨价还价能力;政府实施新能源汽车补贴政策时存在着政府补贴的支付转移,各节点企业一方受益必将带动另一方受益.随着消费者对新能源汽车的偏好增加,政府可以逐渐降低新能源汽车的政府补贴甚至取消新能源汽车的政府补贴政策,政府补贴政策对新能源汽车的发展起着强有力的促进和调节作用.     

绿色供应链中消费者信用支付策略研究

本文考虑了一个由单一供应商和单一电商平台组成的绿色供应链,分析了三种不同支付策略(立即支付策略、分期付款策略以及最低还款策略)对供应链成员最优决策和利润的影响。研究表明:(1)免息期敏感系数对绿色营销水平和供应链成员利润均有正向影响。(2)信用支付会使产品产生溢价现象。但在一定条件下,信用支付策略可同时提高供应链成员利润和绿色营销水平,实现经济绩效和环境绩效的双赢。(3)手续费率较高时,若信用期限系数较小,立即支付策略最优;若信用期限系数适中,分期付款策略最优;否则最低还款策略最优。手续费率较低且信用期较小时,最低还款策略最优,否则立即支付策略最优。在此基础上,针对两种信用支付模型设计了成本分担和收益共享的组合契约,实现了供应链协调。     

Dynamic bidding strategies in search-based advertising

Search-based advertising allows the advertisers to run special campaigns targeted to different groups of potential consumers at low costs. Google, Yahoo and Microsoft advertising programs allow the advertisers to bid for an ad position on the result page of a user’s query when the user searches for a keyword that the advertiser relates to its products or services. The expected revenue generated by the ad depends on the ad position, and the ad positions of the advertisers are concurrently determined after an instantaneous auction based on the bids of the advertisers. The advertisers are charged only when their ads are clicked by the users. To avoid excessive ad expenditures due to sudden surges in the keyword-search activities, each advertiser reserves a fixed finite daily budget, and the ads are not shown in the remainder of the day when the budget is depleted. Arrival times of keyword-search instances, ad positions, ad selections, and sales generated by the ads are random. Therefore, an advertiser faces a dynamic stochastic total net revenue optimization problem subject to a strict budget constraint. Here we formulate and solve this problem using dynamic programming. We show that there is always an optimal dynamic bidding policy. We describe an iterative numerical approximation algorithm that uniformly converges to the optimal solution at an exponential rate of the number of iterations. We illustrate the algorithm on numerical examples. Because dynamic programing calculations of the optimal bidding policies are computationally demanding, we also propose both static and dynamic alternative bidding policies. We numerically compare the performances of optimal and alternative bidding policies by systematically changing each input parameter. The relative percentage total net revenue losses of the alternative bidding policies increases with the budget loading, but were never more than 3.5 % of maximum expected total net revenue. The best alternative to the optimal bidding policy turned out to be a static greedy bidding policy. Finally, statistical estimation of the model parameters is visited.     

视觉显著性对横幅广告注意及记忆效果的影响研究

目前对横幅广告视觉注意和记忆效果的研究出现了“独特性观”和“盲视观”两种矛盾的观点。本研究从任务驱动的视角出发,引入计算神经科学的视觉注意计算模型,借助眼动追踪技术,探讨不同任务驱动下横幅广告的视觉显著性对消费者的注意和记忆效果的影响。结果表明:在不同任务驱动下,横幅广告的视觉显著性对消费者的注意及记忆效果的影响具有显著差异。具体而言,在浏览任务中,视觉显著性高的横幅广告比视觉显著性低的横幅广告能带来更好的广告注意及记忆效果;在搜索任务中,不同视觉显著性水平的横幅广告对消费者的注意和记忆效果的影响没有显著差异。本研究丰富了计算神经科学在营销领域的运用,为企业选择合适的网络广告策略提供借鉴。     

Narrowcasting of wireless advertising in malls

Wireless devices are personal, and the advertiser can schedule ads to reach the prospect at the proper time and place, which makes wireless advertising an ideal direct marketing tool. Industry associations, fearing a backlash from uncontrolled spam, have been careful to emphasize opt-ins. Yet the market has been more promise than reality. In spite of the possibilities of personalization, one of the major complaints has been that the ads are not relevant. Proper targeting and scheduling of wireless ads can go a long way to alleviate this problem. In a recent paper [De Reyck, B., Degraeve, Z., 2003. Broadcast scheduling for mobile advertising. Operations Research 51(4), 509–517] discuss a real implementation of wireless advertising in a shopping mall in London, and provide an integer programming model to schedule ads to various segments of prospects over a week. This model was found to be very effective in increasing response rates without saturating the devices with too many ads. We show in this paper that De Reyck and Degraeve’s model can be made more effective by utilizing information that is readily available to the advertiser. Our test runs show an improvement in the sum of ad priority value of about 150%, depending on the type of traffic in the mall. We provide two enhanced integer programming models that incrementally use additional contact history information to better scheduling of ads. Results from extensive runs are presented and managerial insights discussed.     

Generalized Nonlinear Variational-Like Inequalities

This study investigates how to dynamically allocate resources with a given budget for advertising through Web portals using keyword-activated banner ads on the Internet. Identifying the factors that affect the potential number of banner ad clickthroughs in each portal, we show that the process of budget allocation between the two types of portals (generic vs specialized) that leads to the largest banner clicksthrough in the long run is an optimal control problem. Using techniques of dynamic programming, we find analytical solutions for the optimal budgeting decisions. Our analysis shows that an advertisers optimal portal budgeting depends nonlinearly on the number of visitors who type the same trigger keyword and the average clicksthrough rates, as well as on the advertiser and ad effectiveness. Further, we find that the maximal number of banner clickthroughs from both portals, at time t, depends on the remaining budget until the end of the planning period. The analytical results have useful managerial insight. One of the interesting features of our solution shows that, while a large visitor base may favor the generic portal, other parameters may affect it unfavorably: e.g., lower clickthrough rates of keyword banners from a more heterogeneous audience. Using a specificaction that is consistent with empirical observations, we show that, in the long run, an advertiser must always spend more ad money at the specialized portal.     

供应链互补产品广告投入和捆绑销售联合决策研究

以一个制造商和一个零售商的供应链为研究对象,制造商生产两种互补产品,零售商可选择分开或捆绑两种销售策略。考虑互补品的负交叉弹性和广告外部性的特点,构建基于溢出效应的需求模型,运用博弈论的方法,求解零售商采取分开和捆绑两种销售策略时的均衡结果。通过比较不同销售策略下的均衡结果及利润关系,探讨在分散式和集中式两种情况下,零售商的最优广告投入和捆绑销售的联合决策问题。最后,通过数值算例,讨论产品互补程度和广告成本系数对决策结果的影响。研究结果表明,无论是在分散式还是集中式决策下,当产品互补程度较高或广告成本系数低时,分开销售是占优策略而广告费用较高;当产品互补程度较低且广告成本系数高时,捆绑销售是占优策略且广告费用较低。     

Multistage multiproduct advertising budgeting

We propose and analyze an effective model for the Multistage Multiproduct Advertising Budgeting problem. This model optimizes the advertising investment for several products, by considering cross elasticities, different sales drivers and the whole planning horizon. We derive a simple procedure to compute the optimal advertising budget and its optimal allocation. The model was tested to plan a realistic advertising campaign. We observed that the multistage approach may significantly increase the advertising profit, compared to the successive application of the single stage approach.     

Competitive Advertising Under Uncertainty: A Stochastic Differential Game Approach

We analyze optimal advertising spending in a duopolistic market where each firm's market share depends on its own and its competitor's advertising decisions, and is also subject to stochastic disturbances. We develop a differential game model of advertising in which the dynamic behavior is based on the Sethi stochastic advertising model and the Lanchester model of combat. Particularly important to note is the morphing of the sales decay term in the Sethi model into decay caused by competitive advertising and noncompetitive churn that acts to equalize market shares in the absence of advertising. We derive closed-loop Nash equilibria for symmetric as well as asymmetric competitors. For all cases, explicit solutions and comparative statics are presented.     

Advertising for a new product introduction: A stochastic approach

We formulate a stochastic extension of the Nerlove and Arrow’s advertising model in order to analyze the problem of a new product introduction. The main idea is to introduce some uncertainty aspects in connection both with the advertising action and the goodwill decay, in order to represent the random consequences of the advertising messages and of the word-of-mouth publicity, respectively. The model is stated in terms of the stochastic optimal control theory and a general study is attempted using the stochastic Maximum Principle. Closed form solutions are obtained under linear quadratic assumptions for the cost and the reward functions. Such optimal policies suggest that the decision-maker considers both the above mentioned phenomena as opportunities to increase her/his final reward. After stating some general features of the optimal solutions, we analyze in detail three extreme cases, namely the deterministic model and the stochastic models with either the word-of-mouth effect only, or the lure/repulsion effect only. The optimal policies provide us with some insight on the general effects of the advertising action. Supported by MIUR and University of Padua.     

Innovation diffusion uncertainty,advertising and pricing policies

We develop and analyze a normative and structurally stochastic model of innovation diffusion by depicting the market at an aggregate level. Model dynamics are defined through the flow pattern of individuals that move from the innovation unaware stage, to the innovation aware, and ultimately to the adopter stages. The stochastic evolution of this stage-wise transition unfolds according to tractable stochastic processes and is influenced by such factors as price, word of mouth, and advertisement efforts. In this environment, techniques of contingent claims analysis and stochastic control theory are employed to obtain optimal pricing or advertising policies that maximize the value of the innovation. To account for their optimal adjustment over time, these policies are modeled as positive real-valued adapted processes. Given this setting, policy adjustments over time (i.e. advertising or pricing) are viewed as a value additive sequence of nested real options. We present closed-form analytic results regarding the optimal policies. Simulations provide a numeric insight to the models' behavior.