ACM '59最新文献

{"title":"On pre-conditioning matrices","authors":"E. E. Osborne","doi":"10.1145/612201.612238","DOIUrl":"https://doi.org/10.1145/612201.612238","url":null,"abstract":"That which follows is concerned with improving the condition of a matrix for the purpose of obtaining its eigenvalues and eigenvectors° Some of the difficulties encountered in the eigenvalue problem are essentially due to the fact that the eigenvalues of the given matrix are small compared to the norm of the matrix. Thus it appears that there is something to be gained by applying similarity transformations to the matrix so as to reduce its norm. The notation employed in Householder Ss rl] paper is used here° Thus~ let th A be an n order matrix with complex elements and ~.i(A) (i=i~2,.o .,n) its eigenvalues. Use is made of the Euclidean vector norm i=l i x i I 2 (i) Two matrix norms are considered. One is the Euclidean norm The other norm is the spectral norm (3) where p(A) is the spectral radius 0(A) : max I ki(A) I • (4) i Because of the relation i/~IIAII E~IIAI s~ IIAilE ' (5) a substantial reduction of either norm will involve a reduction of the other.","PeriodicalId":109454,"journal":{"name":"ACM '59","volume":"21 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1959-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129933899","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A new algorithm for algebraic translation","authors":"P. Z. Ingerman","doi":"10.1145/612201.612229","DOIUrl":"https://doi.org/10.1145/612201.612229","url":null,"abstract":"Fo Z~ Ir~german The Computer Center Uni~verslty of Per~usylvanla A new, machine-independent technique for producing a set of controi~. commamd~object triplets from a \"normal '~ algebraic expression has been developed at the University of Penmsylv&mia Computer Center° Input to which the techmiq~e is applied consists of slightly modified (edited) algebraic expressions~ Output after application of the technique takes the form of the triplets, which are considered by some authorities to be mere basic than the three-address intermediate forms found in some cempi~ lets for commercial computerso Each triplet consists of (1) the designation of an object, (2) the command which is to apply to the object and (3) the control point through which the command is exercised. Because this triple level of information appears in a brief form, several advantages accrue. First, many common subexpressions may be eliminated, techuiques such as the '~Russian\" method may be used to minimize temporary storage needs, and expressions which are permutations of other expressions may be evaluated unambiguously. Second, transitions to specific machine codes may be accomplished easily. In fact, the triplets themselves could serve as bhe micro-program for a specific machine. The triplet list which results from application of the technique may best be described as a step-by-step analysis of what must be done to evaluate the original algebraic expression. The technique which produces the triplets embodies several important features. (I) Only one backward scan of the original expression is required during the entire transformation from algebraic notation to triplet notation.","PeriodicalId":109454,"journal":{"name":"ACM '59","volume":"63 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1959-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129558338","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A multi-level code processor","authors":"A. Evans, A. Perlis","doi":"10.1145/612201.612231","DOIUrl":"https://doi.org/10.1145/612201.612231","url":null,"abstract":"Languages used for describing procedures for the so!~ utlon of problems on digital computers have progressed from machinate coae, through symbol±c assemblers and algebraic tr~mslators~ to sj~u~, bol manipulating languages~ In most cases~ in a given procedure dss~ cription, free transition between these forms is not easily accomplish~ ed~ In our opinion this is a serious shortcoming of such systems° T~o basic principles have guided the design of this","PeriodicalId":109454,"journal":{"name":"ACM '59","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1959-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123243688","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A mathematical model for problem queuing in a computer system","authors":"J. L. Bricker","doi":"10.1145/612201.612214","DOIUrl":"https://doi.org/10.1145/612201.612214","url":null,"abstract":"We first compute the steady-state probability, PM' that a machine is operating: average uptime P = M average upt~ + average ~e Employing the technique of differential-difference equations we can then obtain the steady state probability,P~, that a machine is working on a problem of class A. given that it is working on some problem, i Knowing the values of the PS~ one then may determine the probability i that there will be a departure from a computer within the time interval (t, t+ ~i t), given that a problem was on the machine at time t. With this information the state equations of the process are derivable and by passing to the limit we then derive the throughput time probability distribution of the i-th problem class. Thus if w i (t < \"ri) is the probability that an arrival of class A. (in I the steady state) will depart from the system in time < Ti, then L)/. i In the above P (> O) = the probability that a problem arrivlng steady state will have to do any waiting until it enters service~","PeriodicalId":109454,"journal":{"name":"ACM '59","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1959-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124846858","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Machine perception of printed and handwritten forms by means of procedures for assessing and recognizing gestalts","authors":"L. Uhr","doi":"10.1145/612201.612226","DOIUrl":"https://doi.org/10.1145/612201.612226","url":null,"abstract":"The recognition of \"things\" and, more particularly~ the perception of forms~ depends upon the adequacy of the characterization of the interrelations of elements and sub-elements withathe forms themselves. Digital methods can easily eliminate much of this important relational information, if they dessicate the representation of a form into separate atomic cells or spots. But a digital computer can be used to process forms according to a procedure that recognizes successively higher order relations between successively larger elements, or sub-wholes, of a form. It would seem that this type of assessment of a form mirrors, in some relatively undefined sense~ the \"meaning ~ that inheres in the form, and mimics to a certain extent the human perceiver's processing procedures~ at both neurological and psychological levels. The specific problem posed for the machine is the recognition of printed and handwritten language with the success achieved by the human reader. This is a relatively simple problem in comparison to what any ordinary human handles when he recognizes the enormously large number (lO 6 ?) of complex and complexly transformed objects, words, concepts, faces, sounds~ tones of voices emotional implications, etc. in his experience (the broader problem of perception). On the other hand, this is a task difficult enough to be impossible for any existing~ and probably for any proposed, machine.","PeriodicalId":109454,"journal":{"name":"ACM '59","volume":"69 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1959-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128641102","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A computer program for analysis and design of power supply circuitry","authors":"John H. Beaudette","doi":"10.1145/612201.612208","DOIUrl":"https://doi.org/10.1145/612201.612208","url":null,"abstract":"Although many equivalent circuits for any given power supply circuit exist, an attempt has been made to analyze a circuit that includes all the parameters that have any significant contribution to the final analysis.","PeriodicalId":109454,"journal":{"name":"ACM '59","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1959-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129767952","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A non-linear programming algorithm with application to product allocation","authors":"Y. Bard","doi":"10.1145/612201.612234","DOIUrl":"https://doi.org/10.1145/612201.612234","url":null,"abstract":"The algorithm presented here, although applicable to any non-linear programming problem, is designed specifically for efficiency in cases where: (1) the constraints are linear; (2) the function to be maximized is very difficult (time consuming) to compute. Thus, every effort is made to calculate the actual value of the function as rarely as possible; (3) the maximum may be an interior point of the admissible region. To achieve the second objective, we do not travel along the gradient of the function until we hit the nearest constraint; rather, we replace the function by a linear approximation throughout the admissible region, and at once proceed to the maximum of the linear function. A braking procedure ensures that in the vicinity of the maximum the step size diminishes from one iteration to the next.The algorithm was designed primarily for use in a product allocation problem, which satisfies the above conditions.","PeriodicalId":109454,"journal":{"name":"ACM '59","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1959-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129770429","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Job shop simulation on the IBM 704","authors":"Elizabeth B. Ware","doi":"10.1145/612201.612269","DOIUrl":"https://doi.org/10.1145/612201.612269","url":null,"abstract":"The Job Shop Simulator is a general purpose IBM 704 program designed to enable companies with job shop type activities to test out new ideas about their shop operations on a purely c~puter simulated basis~ thus avoiding ex~ pensive and time consuming experimentations with their own ma~Ipower and plant facilities° For examples one might wish to determine the effects on a particular shop brought about by such factors as: i) ii) iii) iv) v) Using new shop dispatch rules; that is~ rules for assigning eligible shop ordersto available machine tools~ Using new scheduling proeedures~ that is; methods for de-terming customer order due dates and theoretical, mac~dne operation start dates° Changing machine tool and manpower capacities~ including additional working shifts° Using new customer order configurations° Using generated statistical fluctuations in machining~ setups and transit times° Further~or% we might wish to measure the effects of the above in terms of possible improved conditions such as: i) ii) ill) The reduction of in-process inventory cost. The reduction of late customer order completions° The reduction of machine tool and labor idleness° The Job Shop Simulator encompasses all of these factors and more to enable each company to answer these questions by experimentating with its own individual shop on a simulated basis° By repeated simulations under studied changes in the shop variables, there should evolve the best shop conditions for the company's particular management requirements° Of prime importance to all Manufacturing Control firms is the problem of predicting changes in customer order completion dates and in-process in~ ventory costs due to changes in the number of existing machine tools and machine tool operators. Quite frequently one may feel that a few more machine tools or a few more machine tool operators will materially improve the late order situation. The Job Shop Simulator enables the user to determine these effects° By experimentation with different numbers of machine tool and machine tool operator combinations~ one can by successive simulations determine the influences of these factors on the shop. Before the Job Shop Simulator can begin its simulation a set of customer orders must be made available to it to work upon. To the simulator each order is a prescribed sequence of machine tool operations with associated setup and machine times, including a customer order due date and an initial raw material cost. (One can choose to let the simulator determine the customer order due date). The simulator …","PeriodicalId":109454,"journal":{"name":"ACM '59","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1959-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124044528","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"LISP: a programming system for symbolic manipulations","authors":"J. McCarthy","doi":"10.1145/612201.612243","DOIUrl":"https://doi.org/10.1145/612201.612243","url":null,"abstract":"LISP (for LISt Processor) is a programming system for the IBM 704 being developed by the Artificial Intelligence Group at MIT. We are developing it in order to program the Advice Taker which is to be a system for instructing a machine in a combination of declarative and imperative sentences.","PeriodicalId":109454,"journal":{"name":"ACM '59","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1959-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115109256","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Probabilistic indexing: a statistical approach to the library problem","authors":"M. Maron, J. L. Kuhns, L. C. Ray","doi":"10.1145/612201.612218","DOIUrl":"https://doi.org/10.1145/612201.612218","url":null,"abstract":"This paper describes a novel and promising approach to the problem of automatic literature searching and retrieval. The technique, called \"Probabilistic Indexing\", can be applied to any type of library indexing system whether it be classification indexing, coordinate indexing, etc., and the technique is directly amenable to automatic handling by digital computing equipment.","PeriodicalId":109454,"journal":{"name":"ACM '59","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1959-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124632433","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}